Techstrong TV August 8, 2025

Hey everyone, it's Alan Shimmel for Techstrong. As I said in the opening, if you've never been a blackout or you're not here this year and you're wondering what it was like, you've come to the right text. Strong gang.

This is a very special episode, excuse me, won't be joined by my usual gang compadres. It's, this is solo, but I've been out here this week interviewing and talking and catching up, of course, on all my friends in the cybersecurity space. And I gotta tell you, between RSA and Black Hat, it's always great to come out and see people that I know, some cases 20, 25 and more years, and, uh, to hear what's going on in the security world.

We've got some great video we're gonna be playing over the course of next week from our Black Hat interviews, but I wanted to bring three special ones to you today. First one is an exclusive Textron. It is by the Black Hat Knock Crew themselves, and it's an exclusive behind the scenes look at what goes on in the Black Hat Knock, perhaps one of the most heavily defended and heavily attacked knocks, uh, in the private world because everyone would love to scalp say they broke Black Hat's Network.

But, uh, it's a great episode, so you'll check that out. I'm gonna follow that up with a great conversation we had with our friends, uh, from Palo Alto Networks about the Cortex Cloud offering and what's new there. Like everything else at Black Hat, there's a little AI in there for you too.

And then finally, we're gonna end with a, uh, a video shot here in our studio, rather than on the floor of a interview with my good friend, rich Mogul of Fireman. Of course, many of you know Richman's years at Garner and Securosis and Disrupt ops. And, and, you know, rich is one of the smartest people in security.

I know he wrote the Cloud Security Alliances classes, but he's here at, uh, he's been at Fireman now for a number of years, and he's talking about network security, N-S-P-M-A recent Survey Fireman did on that, as well as an, uh, recent announcement partnering with Illumio on, on, uh, microsegmentation and Zero Trust. So three good ones, and if you like these, there's a lot more behind them. Stay tuned.

Next week, we are gonna be playing all of our great Black hat videos, so thanks for joining us today on this special Techron Gang. Let's go ahead over now and find out about the Black Hat Knock. Hi, I'm James Pope, and welcome to the Black Hat Knock.

I am the SOC leader here for the Black Hat Knock, and I'm also the Technical Marketing Engineering Director for Core Light. We start prepping a long time before this conference to get ready to make sure that we can build an entire network from scratch. We bring in the ISP, we bring in switching firewalls access points, and then we bring in all of our security tools on top to make sure that A is available and b, that it's secure.

Part of the security part is we have core light doing full P cap and network visibility of all the traffic that is here, or threat hunting, finding the really bad and the bad. We call these blackout positives. Those are things where they are a legit bad thing, but we're gonna let it happen on this network.

A student comes to learn about some how to run a malicious tool or carry out an attack, and they get taught that, and we see that across the wire, and we let that happen in this environment. We care about the things that are truly bad and that bad students attacking students, somebody attacking registration or backbone. So we have a lot of eyes on a lot of screens paying attention to make sure that we are spotting those things.

We're alerting on it, and we're res responding appropriately to that. Some of the findings and things that we find every time is users who have a green checkbox in the bottom corner, so everything's secure, but they have a VPN that's leaking out credentials. They got some sassy tool that is sending all the proxy of all their information out in Clear Text.

Uh, I'm calling it the rise of, uh, ai, you know, everybody's calling it that. But the rise of Vibes, vibe Coding is taking off, and we're seeing a lot more apps, whether that be a weather app or whether that be, uh, leaking out all the GPS information or a, this year, a few different chat applications that are sending out all the corresponding information of that entire chat, including their voice and translation. So we're just seeing way more stuff in clear than we would like to see, especially at a security conference or let alone from any of these corporate laptops in this environment.

We do that with a lot of different ways. We use detections, search based alerts. We have Yara signatures.

We're leveraging Zeke and CTA on the back end. And then we're using ML hits and also AI detections. We work with our partners.

We do truly call them partners here. No, no tool out there can decide. It wants to be a part of the Black Hat Knock.

We go and choose the best of the best, and we ask them to come, bring their tools, bring their people, and, uh, make sure that we have a good experience. Those partners are Cisco, Palo Alto Networks, Arista, and Core Light. This year, we're leveraging a lot more with ai.

We're using Palo Alto Networks XIM to do a lot of summarization and categorization. We're also using a core light MCP server that lets us directly from an LLM client, whether it be Gemini or CLO or anything else, or a a Slack bot where you can ask it, tell me about this incident. Tell me about this IP address, MAC Address, FQDN.

And we're leveraging that MCP server to go into the raw data, give us a relevant pieces of information and bring them back. It's working amazingly well for tier one, tier two people who might not know how to write SXQL queries or SPL queries, or insert some version of QL queries. They can ask a question, get the results, and then they can start acting on that information, uh, quicker and faster.

Uh, this year we also had two different organizations. One was a bank, one was a Fortune 50, who their security tools were actually giving away information about their machine, their patch level, their logs through misconfiguration, or just not enabling TLS. So we are in the year of vibing, but vibe and verify, validate that the things that you built are good.

I have the luxury of having all these expensive tools where I can look at it and validate that my stuff is good, but Zeke is free. Scot is free. TCP dump is free, Wireshark is free.

Go and validate that your stuff is not leaking out things before you send them to conferences or even just to go to your coffee shop or fast food where they're on any hotel network, and also leaking out that same information. While we do have a very highly customized network here that's very purpose built for this conference, there's a lot of things that people can do in their organizations they can take from this and use at their orgs. One is all the detections and alerts we see.

If a, somebody gets up and does a presentation on a brand new thing that they find, inevitably somebody turns around and tries to do that. So we want to look for that. We create detections for that, and then we build those integrations with other partners and those detections that help our customers going forward.

But yeah, organization, you wanna make sure that your stuff is secure. Your people are secure on their end points, not just for Black Hat, but for all conferences everywhere, where they're operating that they're doing in a secure manner. Thanks for coming and visiting us in the Black Hat Knock.

And we are here for the US Show Europe and Asia. You wanna learn more? We do have a Twitch stream.

You can watch us live in our fishbowl, but if you come to the conferences, you can come in and do a tour and see what's happening in here. Come learn more about Black Hat Knock and come learn more about Coral Light. Hey everyone, welcome back here to Techstrong tv.

So we came off of that crazy show floor to our luxurious broadcast suite here at the Luxor Hotel. MGM Tell my Wife I love her. Yes.

Um, but thanks for joining us in our continuing Black Hat 2025 coverage. My next guest really needs no introduction to, to security people and, and, uh, our audience at Techstrong. It's my friend Rich Mogul.

First of all, rich, welcome back to Techstrong tv. Thanks. Thanks for only the best for you, rich.

Only the best. But, um, thanks for coming up and being with us. I appreciate it.

Rich. Of course, you're at Fireman. Yep.

You know, I forgot your title. Is it VP of Cloud Security? S VP of Cloud Security.

You got it. SP of Cloud security. The S is for special.

Well, you know, I wrote an article last month. The S in Vibe, in Vibe coding stands for security. And you know, that brings me up.

Remember we did a podcast once with the CEO of Mongo. DI Will never forget that. Me, I bring it up all the time And I talk about it all The time.

And why not is that, is no sequel, mean no security, and they said, we'll have security when our customers Was their answer. And then everybody got breached and then they added security. And I think the same thing's gonna happen with Vibe.

Yep. Agree. When people start demanding security, they'll do something about it.

But until then, as I said, the s in right. Coating stands for security. Um, but Rich, you're at Fireman as we mentioned, but of course, if you know Rich Long distinguished career as a Gartner analyst covering the, uh, Data security D-L-P-D-L-P Space.

Yes. Yep. That, that, well, those were my days.

Those were my years. DLP and stuff like that. And then of course, rich and our good friend Mike Rothman went on to found, uh, Securosis.

Yep. Uh, kind of reset, broke the mold in security analysts firms over the years. And then you guys, rich, you were the primary driver of, of a product vision that that came out and that's how you came to Firemont.

Yeah. So they, uh, acquired our startup Disrupt ops about mm-hmm. Three or so years ago.

And, uh, yeah. And then so It's been a ride. It's been a ride, my friend.

It's, Uh, yeah. Any little corner of this industry you could hit. I've probably, I'm, well, you know, I always like to think it's a round room and there are no corners.

But you're, you're right. We've been there. But you, you wanna know the nice thing coming to Black Hat?

Our next guest is in the green room waiting for us here. Fred Wilmut. I've had a chance to meet so many people and you've met more, you know, more than me.

And, but we've met so many people and you come to this or you come to an RSA, maybe twice a year, we get together and, uh, it's good to see these people. I mean, some of these relationships are 20, 25 or more years old. I've Known you for over 20 years.

Absolutely. I'm ashamed to tell you longer than that, my friend, because I think the first security bloggers network was over 20 years ago. Party.

Yeah. I think it was 2003. I, I'm bad at math.

No, I know. Well, I, it's easy 'cause we're in a 25 year, so it's easy to say what 25 years is, right? Yeah.

But next year it'll throw me off. Anyway. Hey, rich, we're here to talk a little bit about Fireman, though.

I think most of our audience knows Fireman, but for those who maybe aren't, why don't we start there at that 50,000 foot level? What, what is Fire? Yeah.

Fireman focuses on security operations, and the area that we're most focused on is network security policy management. So NSPM is the core product. Uh, we do also have, uh, my old product, which is a cloud security posture management product.

Uh, we have an asset manager product as well. Okay. And if you have really large, complex, uh, it doesn't even need to be really large.

If you need to manage firewalls from different vendors, different environments, make sure those things are all compliant, uh, fireman is kind of the best at that. Yep. And just, you know, to serve as, uh, cybersecurity historian Fireman, of course, was spun out of Gary Fish's.

Yep. Fishnet Security. The CTO of Fishnet was a guy named Jody Brazil.

Yep. Brazel. And, and Jody, they spun it out as fireman.

And Jody was the first CEO he left for a while, but he came back. He's still CEO. Yeah.

So it was, uh, it was a pretty wild story. So Jody invented it, basically because he was doing these consulting projects, and he did the, let's see if I can automate myself out of a job. Mm-hmm.

And he came up with ways to do automation, connected to all these different firewalls and have this consistent policy enforcement went to Gary. Gary spun it out. So Jody was, he was the tech founder.

Everything became CEO. Now, when he left, after some, I guess some external investment years later, uh, I was his next startup. So he was my co-founder of Disrupt ops, uh, him, Brandy Peterson, Mike Roth, and Adrian Lane.

We all founded this company, disrupt Ops. And then Fireman acquired Disrupt Ops, and it was like a reverse merger because Jodi then took fireman back over again. Right.

It's an, it is an interesting story, but, you know, politics makes strange bed flows. Yeah. And security stories are constantly, you know, strange.

It's a strange engine. And circular. And circular.

Right. Exactly. No corners.

Um, but Rich, I, I, you know, speaking of network policy management and I, I left a a, an A, uh, an initial outta there, didn't I? It's NSPN Network Network Security. Policy Management Management, yep.

Fireman recently put out a report. Yep. Talk to us.

What's it about? So, Uh, we had this new product called Insights. Mm-hmm.

And well, you know, kind of product kind of feature. So we actually leveraged some of the stuff that I had done in cloud as the base platform for this, or me, our team. I mean, it was 30 people won't get acquired.

But, uh, the insights product for customers that are willing to share the, uh, um, use this, it uses their data and does analysis to help them optimize their use of their firewalls. So it gives you all this really wild reporting and stuff that nobody else has seen before. Well, we found out that there was, uh, some interesting things that we didn't even know, because historically we've got our little silos of customers here, here, and here.

And we had a way to look at kind of the data in the big picture. Now again, all privacy preserving customer driven, like, let's, let's be careful we're not stealing our customer's data, but we found that like 90% of firewalls had, uh, critical policy failures. And what, what do we mean by that is it's a compliance failure, uh, an obvious compliance failure.

And it can be anything like somebody left Port 22 open word that shouldn't have been, or, uh, clearex protocols where it shouldn't have been or, or anything along those lines. So those policies, and, and there are standards around, like PCI, for example, we map those specific firewall rules to what PCI requires. And there were that the high degree of failure, but then there's some, or sorry, it was 60%.

I'm gonna cheat and pull my numbers up. 60%. Okay.

The high severity compliance checks, the 90% is actually 95% of numbers, uh, falling Short of critical levels. Yeah. Well, it wasn't even that.

It's well inefficiencies, 95% of the application objects that people define. So you can define application objects and firewall rules were used. Right.

So you're turning on your burden CPU cycles. You have these bigger, complex policies that are gonna be problematic to deal with. And, uh, and You're not in compliance, that you're not secure.

Yeah. So here's what I find not fascinating, revolting, that, you know, I've known about fireman since he spun it out. Yeah.

I remember going to Kansas City, talking to them, um, And We've had firewalls, next generation, firewalls, web application firewalls, this firewall, that firewall. We've had companies like Fireman and, and some of their competitors back in the day, TwoFin and, uh, I forgot the other one, I forgot 'em all, but Whoever they are, but, you know, that have preached firewall policy management religiously for 15, 20 years. Well, It's in every audit and every assessment.

So why, why do we still deal with this? Why are we still It's ai Help me. Yeah.

I mean, can AI automate this once and for all? I Mean, and we actually have some of that available in insights to help you, like, explore your environment. So we have an AI chat bot up there, uh, which you didn't even know when you asked me the question, but the, it's more of, um, so this was new to me.

Like even though I've been in security forever, I haven't really dug into firewalls too much. And, uh, after the acquisition, even though I'm very cloud focused, uh, some of what I had to do also began having to focus a lot more on the network security angle. Specifically.

There's so many reasons why. One is like somebody will put a rule in to get something working. Mm-hmm.

They'll forget to take it out or manually trying to manage these rules in these heterogeneous environments. If you have, you know, checkpoint IMP Palo and Cisco and Fortinet, and a lot of organizations do, and even if they try to standardize on one, then they're gonna acquire or have a merger or something like that, and they're gonna get other ones out there. So it just creates all of these extra levels of complexity.

The other is, is when you're dealing with these at scale, the process of manning managing those rule changes and pushing those out to where they need to be, uh, it blew me away how much goes into that. There's organizations that literally have dozens of people dedicated to just managing firewall world changes. And it's not an exaggeration.

I, I was like, wait, you have how many people? And I'm like, don't you have any automation? They go, yes, this is after the automation.

These are all the exceptions. 'cause some of these orgs just have these, you know, incredibly large, complex environments. Oh, Absolutely.

And then the mid-size, they don't have enough people to manage what they do have. And that's also been a problem. Yeah.

Forever and ever. Right. But that's why we love the insights, because that is exposing information to them.

That was, that data was al always there. But within the, the market, like, we weren't providing that in a way that was like impactful. Like, you can go to your CEO go, we're failing 60%.

I mean, that's the average in the report, not the 90 I said at first. Right. The 60% we're failing 60% of our compliance checks, you know, that are higher above.

Mm-hmm. We're, we have 95% of our application objects aren't even used. Like, that's just wasted space and added complexity.

Yep. So that's the kind of stuff that was like the, I'll, I'll be honest, when our team saw the results, they were like, oh, this is really good. Well, it's good for fireman, right.

But Well, yeah, but it's bad for what's going on out there. I think you pulled the 90% number, 60% of enterprise enterprise firewalls fail high severity compliance checks. Yep.

Another 34% falling short at critical levels. Yeah. So that's where you probably got 90, 90%, 94%.

I wanna pivot if we can a little bit. Recently Fireman announced an integration with Illumio. Yep.

The Zero Trust. And of course, Ilum Illumio is the leader in the segment network segmentation market. Let's talk about that.

Yeah. So, and that was, uh, actually what one of the things that I was involved with. So that was, uh, kind of the products that I work on with the Illumio integration.

So we're not releasing all the specific technical details around this, but when you're using these micro-segmentation products and you have traditional firewalls and other network security controls in your environment, uh, there can be conflict. So a lot of times the reason an enterprise is gonna bring in Illumio is because of, uh, a couple of different things. Maybe not enough firewalls, or they need deeper segmentation, you know, and there's cost effectiveness becomes a factor there.

Uh, you can't necessarily drop boxes everywhere in. And then there's also the additional layer of what products like Lumira are good for is they start giving you a better ability to manage rules based on what something is, as opposed to firewalls, which were designed purely to protect a good network from a bad network. Well, the problem that you can encounter is that for products like Lumio at work, they have to have agents everywhere.

And so there's a couple of different layers of issues where you, you can potentially run into issues. One of those is, uh, imagine you are a hospital or manufacturing or other facilities. You can't always install agents on everything.

Mm-hmm. And so you're still gonna need the firewalls to provide the rules, uh, around protecting those objects. But you still want it to work well with Illumio.

So what we've done a lot of the, and as we announced more about this, get out more details, but it actually can glue together the firewalls and illumio in intelligent ways so that they can actually be more compatible. The other issue is, is what if you want your, uh, illumio assets to talk to each other, but you've gotta get across the firewalls. And sometimes that can be a problem as well.

Sure. So those are like the two most common problems that we've kind of built this to, uh, go ahead and be able to address. And, and that's why it's great 'cause we can get to the asset level, attribute level security, and we can do it with your existing firewalls and then, and have that also work with the microsegmentation with the rail.

Got it. Now look, it's a zero trust play. Yeah.

But we should also mention it. It is, uh, it's, it's about resilience too. Yep.

Right. And, and that's a big thing, right. You know, people may not associate, uh, network security, uh, posture manager NSPM with resilience, but that's part of the resilience model, right.

Is try to contain Yeah. Where we, where we, where we're threatened, where something goes on, right. So we don't lose the whole ship.

Yeah. Being able to respond more dynamically. So there's that security, resilience play, and then there's also the resilience of what if the firewall goes down or this goes down or that goes down, being able to actually, you know, have the ability to like, update your environments to account for those kinds of situations.

Yeah. And, and it plays into the zero trust thing, which I, I think is finally, you know, with all due respect to John Kinder, that guy I was talking to John a couple weeks ago, a lot of people poo-pooed it and gave it a hard time, but it's really become part of the Concept. Every, every company I talk, like I had to do a bunch of research for our new products that we're working on.

And, uh, it blew me away that they all had some kind of zero trust initiative. Yeah. It's, it's the way it is.

Yeah. It's the way it is. Anyway, rich, I think we covered the topics that our corporate overlords have, uh, asked us to, to cover.

Is there anything else that we missed, you think, or? No, it was, uh, I mean, pretty good. The, uh, you know, tying in a little bit back to the zero trust piece of it too.

The part is is I like, like you, I poo-pooed some of the early stuff. Mm-hmm. Let's, let's be honest, we all Did.

Yeah. And, but I've come around on it because, uh, particularly now, because we have all this complexity, uh, that's been added to our networks with cloud and with containers and, you know, ephemeral, virtualized assets and everything else. And like a lot of our security models just haven't worked for that on the network security side because it's port protocol source destination.

And as somebody who's very cloud centric, this has been the, I had forgotten how much harder a problem. It's in a data like cloud. I have a lot of capabilities.

I can do all these. You start, and that's funny. 'cause initially we thought we didn't have that in the cloud.

Right? We didn't have enough control, we didn't have enough insight, we didn't have enough ability to manipulate what we needed. But now you're saying, you know, I'm so used to doing that, that this stuff in the data center is a lot harder.

It is a lot harder. But some of those principles, like in cloud, I can very easily write rules that refer to the assets or the attributes. I mean, that's a really powerful part of this.

Mm-hmm. Like this asset with these tags connect to this thing over here. And those are things that we have really struggled intensely with in the data center.

And so, you know, either with our, you know, bringing that asset intelligence and doing it in a way that works for enterprises, like that's a big part of all of this is, is really easy to show this stuff off in a lab. Agreed. But you go into some of these large, it's A real world.

And our clients are huge. Some of these environments. Oh, I, I remember that.

I mean, I, you know, I know the firewall story. What, what freaked me out when I first became from really familiar with Fireman is you had customers who had dozens, if not hundreds of firewalls. Hundreds or thousands is not uncommon.

Yeah. It's crazy that have to be managed and now in multiple locations. And now you've gotta layer in cloud capabilities, like understand the cloud network and then harmonize the cloud network with the on-premises network.

Um, because you've got all this hybrid stuff that needs to talk to each other. And yet in the end, we want this thing to talk to this thing and not talk to, it's A relatively simple thing, right? Yeah.

So that's where like this lumio partnership and other things that, you know, come out someday be being able to have more of an ability to kind of make those decisions, uh, and have that, that higher level intelligence so you're not down to a five couple of firewall rule anymore. I Get it. Hey Rich, we're about to add a time.

I appreciate you coming up here to the thanks for having Taj Mahal and, uh, Am I allowed to leave? No, you just, you gotta see the table. Why are plastic?

Yeah. The corner there? Yeah.

We're gonna edit all this out, guys. Um, just make sure you stop in the bathroom. Wash your hands real good.

Okay. Rich Mogul Fireman here at Black Hat. We're gonna take a break.

Hey everyone. We're back here at Black Hat on the show floor. If you can't tell from the lights and the noise, it's live.

Uh, people are all over. But we carved out some room here at the Palo Alto, uh, exhibit, if we could call it that booth. If you want.

Let me introduce you to Orion Ca. Caseta Cato. Cato.

Yep. I apologize, Orion. It's all good.

I did get Orion right, though. You got the important part, like the star. But, um, Orion, thanks for joining us here on Techstrong tv.

Say hello to our audience, give them a little bit about you. Yeah. Um, my name's Orion Cato.

I'm Senior Director of product marketing for the Cortex Cloud Solutions of Palo Alto Networks. Um, I've been kicking around the cybersecurity space for about 20 years. About a third of my career is in application security, a third in security operations, and a third in cloud security.

It's a nice, it's a nice third, third and a third kind of rounded out thing. Um, Orion look, our audience knows Palo Alto. Yep.

But you know, this isn't like your grandma's Palo Alto. It's, it's more than just next Gen firewall, obviously you mentioned Cortex Cloud. Yep.

That's the cloud security offering. There's Unit 42, the, uh, the, uh, security research. Yep.

Division. You guys have done a couple of acquisitions recently. Some bigger than others.

They're not all closed yet, but you could follow along on Security Boulevard to get all the news on that. Uh, what have I left out? Uh, the thing we're talking about today is Cortex Cloud, which you talked about a minute ago.

Um, what Cortex Cloud is, is essentially a re-architecture of our Prisma Cloud solutions, which is our cloud security, uh, portfolio. What we essentially did was we took, you know, our well-known market leading solutions, Prisma Cloud, we re-architected them on top of our security operations platform, which is Cortex. That's how we got to Cortex Cloud.

The benefit though, is essentially we now have, you know, our application security, our cloud posture security, our cloud runtime security solutions, all on the same unified data layer layer as our security operations solutions. So they have, you know, context running all the way from code to cloud, and then a unified policy and AI engine and the automation engine of Cortex. I love it.

Um, look, you, you shoved a lot there in a little bit of time. I did. I hope I can, uh, you know, distill it well, We're not gonna play it any slower because that's not gonna help.

But let's peel the lead the onion back a couple little, let's dive into each of these things. Sure. So the problem in the cloud security market essentially is fragmentation and the natural silos that exist within the platform.

So cloud security essentially is, um, three things. It's your application security, so the app, the applications you build, there's a whole bunch of six or seven tools that everybody has to have to make that work. Then you have your cloud posture security, which is essentially kind of a, a proactive battening down of the hatches and making sure your misconfigurations are dealt with.

And then finally you have your runtime security. So, you know, finding attacks in real time reacting to those. And so what Cortex Cloud is, is essentially taking the, you know, 16 or so solutions that an average company would have to build these together, and then putting 'em onto one platform to break down the silos of those tools and the silos of the teams that run those, which is application security, uh, your, uh, cloud security team, and then your security operations team all in one platform.

I love it. I love it. Look, the theme of this year's show is definitely, uh, ai.

Yeah. And I know you guys are doing stuff with AI within Cortex. Talk to us about that a little bit.

Yeah, I think it really has, um, two implications on, has two implications on our, uh, portfolio. The first is, people don't often think about this, but when you're building an AI application, it's actually a cloud native application underneath it. So the more AI you use, the more cloud you use.

And our solutions help you to secure that cloud. So that's the first thing. The second thing is, you know, the proliferation of AI is starting to reach all sorts of areas in one that's very important is AI generated code.

So about 95% of code will be AI generated by 2030 according to Microsoft Report. Um, what that means though is, you know, we have a lot of code being generated by ai and a lot of that is vulnerable and being pushed into production to the point that around 74% of cloud breaches are now due to insecure go. So really drives home the need for not only greater cloud security, but also, you know, an integrated approach that brings application security into that.

I love it. Very cool. Um, let's talk black hat.

Yeah. What are you hearing from people? What's been the reception for what's happening with Cortex and Palo Alto on, on a broader stage?

Yeah, it's a great show this year. Um, there's a lot of buzz in the air. So we announced our application Security Posture Management, our A STN features.

Yep, yep. Um, it's been very, very well received. Um, you know, at least for this booth, it's one of the hot topics.

I, I think A SPM is pretty hot all around. What about, I, I know you're not part of, but unit 42 news there. Uh, yeah.

We recently released the, uh, I think the 2025 global IR reports. Um, so, you know, lots of good tidbits in that. Lots of good research.

The unit 42 guys are always cooking up amazing stuff. It's a good, it's a talented team they got over there. Absolutely.

Absolutely. How can people engage with Cortex Orion? Yeah.

Um, the best way to get involved, uh, the best way to get involved is to, uh, check out our website. So both the Cortex Solutions and the Cortex Cloud solutions. com.

Um, we also have live interactive demos that you can, you can do for the products and, uh, you know, kick the tires. Fantastic. Hey, I, I know we, we took over the, uh, the theater here.

Yeah. We gotta give it back. But I want to thank you for coming on and getting us a little bit educated.

Our cortex. Of course. It's Palo Alto.

com. ComCom. And then they could click through to Cortex from there.

Yep. Alright, Orion, thank you very much. Enjoy the rest of Black hat, keep up the good work.

Thank you. Cortex Cloud from Palo Alto Networks here at Black Hat. We are, well, we're not live, we're edited, but we hope you've enjoyed this.

Stay tuned for more Black Hat coverage on Techstrong tv. Hey everyone, it's Alan back here. I hope you've enjoyed these three, uh, videos, which give you a taste of what we've been covering here at Black Hat this week.

As I mentioned, next week, we will be playing the entirety of our Black Hat video series. There's some great ones including Kar Quala, CEO, talking about, uh, agentic ai, uh, rocks Risk Operation Centers. We're gonna be talking with, uh, my friends, Jeremiah Grossman and Robert Hansen, our snake about their new venture.

Uh, just a lot more, a lot more push. Security is another one that comes to mind, was really great. I, I think you're gonna enjoy them all.

But for now, that wraps up our text on gang for this Friday. I hope you enjoy it. I'm Alan Shimel.

Have a great day. SaaS is dead and I think that's a fun one because I think SaaS is very dead. We talked to, um, bill McDermott that serves now two weeks ago.

We've talked to a couple people in that space where being a SaaS company is now having the nicest house in the worst neighborhood. Welcome everyone to episode two of season two of Investing with the Boys. We got all three of your boys over here.

Again, we got Logical Shai and we got Sam, myself here for the second episode kickoff. And I'm gonna just jump right into it. It's been a crazy week, two weeks with earnings.

We've seen crazy moves and volatility from both ends. Very popular retail stocks making big moves up and down. The one that really stands out to me out of all of them is, especially with the recent news and the recent moves it's had, I mean the last month the thing was like up 20%, down 20%, whatever is hims and hers.

This thought, you know, it's actually really interesting on the most recent earnings is that it was the first quarterly revenue missed on analyst expectations. But on top of that, it hit basically a new all time high intraday last Monday. And then it went all the way down to 52 bucks.

And then it recovered that. And then it gave that all back after Novo Nortis came out with the news that they opened up litigation cases toward a lot of telehealth companies. I know we have a lot to talk about this one, so let's just get right into it.

Shy, what are your thoughts on their earnings and what are your thoughts on the company moving forward? So if you're a hams and her, uh, investor, it's not for the fan of heart. You gotta know, that's probably the most volatile stock that I track out there.

And this is coming from a Rocket lab, ion Q shareholder, et cetera. I think there's two ways of looking at HIMSS right now, and both are true. On one hand you have the business I think was growing 70 or 75% top line.

I think there're just the EBITDA is expanding at over a hundred percent. They're adding subscribers at scale on a market that kind of has failed incumbents have failed to disrupt, that's like clear business execution. They're doing a great job at that front.

And it's also not coming from these like lost leading pricing, uh, F type stuff, or one time pops like I'm pretty sure the average subscriber, uh, the spending when is like $74 a month. I think that was like 30% increase year over year. So like the platform has proven that it can drive sustained engagement through personalization like that.

That's their whole thing. Uh, it's not going to be just this commoditization prescription delivery system, but this drop is justified because I do believe this is the first real stumble as a public company. And it wasn't just a revenue miss, it was a sequential decline in revenue and margin compression kind of happened at the same time, at a time when the business is supposed to be scaling.

So there is somewhat of a perfect storm of, on paper, it looks a little off the street was pricing in, uh, the transitioning away from the compound GLP ones that would, that would be a lot more smoother than it actually was. And they thought that the core business would be picking up a slack and that hams would be continuing their compounding like they've done like clockwork. But I think this earnings re report proved that it's gonna be a lot messier than anticipated.

So you have to revise the expectations. And the GLP one alone went down 20% quarter to quarter. That's a huge pullback.

And that was their biggest growth engine last year. I know this is not a GLP one business, but it's still a anchor on what the stock is gonna do going forward because it has that transition phase right now. Yes, they reinfor, I think they reaffirmed the reaffirmed their weight loss revenue guide, but now that the target assumes a sharp reacceleration in Q4, I believe, and that is a high bar to, uh, to pass because there is no room for further slippage.

So meanwhile, I think the rest of the bus, uh, business, the sexual health, the dermatology, the mental health, they actually all went flat or decline sequentially as well once you stripped out to GLP one. So there is somewhat of a high expectations and potential problem if they don't hit their full year guide. Uh, and, and the fact that adding in the fact that the average revenue per subscriber also fell $10 a quarter to quarter, despite I think hims positioning themselves in the high a CV part of their platform.

The, essentially, this is a long winded answer saying there is a lot in flux right now, and especially in a highly commoditized like space like telehealth. Like when you have Nova Nortis suing you and you have like this, uh, commodity, your, your industry's a commodity narrative and these are the fundamentals you reported. Like it's kind of a very noisy situation, especially the run up has had this year that it feels like there needs to be a digestion period.

I don't think it's gonna go in the twenties. I'm not saying that I just think it's $50 looks like a very clear floor until they can like iron out what's happening with the business. But I don't know what, what do you guys think?

Do you have any other inputs on that? Uh, what to do with himss? So I mean, $50 is a floor would be today as of our recording on Wednesday that it, it's dropped pretty con pretty considerably and been extremely volatile.

I mean, uh, I I've been invested in him since uh, since the high teens, um, not even for that long, which is really interesting 'cause it's had so many ups and downs and you know, it can be pretty advantageous for a lot of traders in the market. And this is a trader stock. It has very high premiums.

Uh, a lot of people do practice selling cover calls and then so on. But going back to the fundamentals, uh, I think that ever since their relationship or partnership with Novo Nortis did break off and all the allegations that they put out there, you know, which is of course what they're gonna do in their own best interests. And then we heard that Nova Nordisk is opening, I think it was 14 uh, cases with telehealth companies.

They didn't specifically name HIMSS per se, but it was almost like they said, we're gonna sue 14 telehealth companies. We're gonna open 14 cases and they're gonna be targeted toward companies that rhymes with bims and Burrs, right? Like it's obviously for himss.

And it was actually pretty interesting because there's another telehealth company out there, much smaller cap company called Life md, where they do focus more on the telehealth sector. And then there's also a business that they do consume on top of that, which is included in the metrics. They just actually just reported earnings after hours yesterday.

Didn't get to dig into it too much, but the stock is down 45% today. It made its all time high around $15 and it's dropped nearly 58% from that all time high. And this was another telehealth stock that was also running hot in addition to hims.

So I would really attribute this to not just the fact that HIMSS being a growth stock, of course with being an online pharmacy for a lot of, uh, for a lot of generic branding medication, also with the generic branding or compounded glide. Uh, but also the fact that the entire sector in terms of tech when it comes to disruptors in, uh, in med Medicare, not Medicare, but when it comes to disruptors in pharmaceutical products as well as uh, online medical professional assistance, they all got hit. And we all know the whole story with United Health.

There is a bigger narrative going on here that's hitting healthcare stocks all across the board. Whether they're generational, disruptive healthcare companies or where they're legacy conglomerates like United Health and so on. They're all taking a hit across the board.

And I think there's an underlying macro reason for all that happening, which I believe is starting to come to fruition. But the fact that you have leaders in, uh, leaders in, uh, semaglutide out there, of course you had Ozempic and you had Wegovy from Novo Nordis starting to stir up that drama and the whole situation. It does make investors a little bit uneasy in the scenario.

However, I look at the quarter, I'm looking at it with, uh, you know, with probably a little bit more optimistic in that sense, mostly because I am biased in the situation. But I mean, they did grow subscribers by 73,000 quarter over quarter and they had some good numbers as far as their, uh, as far as the segmental metrics. But yeah, I think missing missing that revenue expectation for the first quarter ever was probably a big thing to kind of put like investors and the momentum itself at like a stand hold.

'cause now it's like, hold on a second. Is this like the leader that we are all expecting or are they gonna run into a little bit of a digestive period, which you're saying, which I do think they will. Uh, the thing is, is that as far as the Marcos and as far as we goes as fundamental investors, this is a trader stock man.

Like they can bring this thing all the way down to 30 bucks and bring it right back up to 70 bucks at the very next day. You do need that stomach to own this stock. In fact, I would always attribute that, and I think logical will agree me on with the, with this, is that you need to have the position sizing when it comes to something like this.

There are people who are like a hundred percent invest in their portfolio in himss. Great, you have that conviction, go ahead, do what you gotta do. But for people who wanna diversify portfolio being in the leaders of the entire world, it's so difficult to put that much leverage on this company as far as concentration of portfolio risk goes.

Well, I I also want to add, this isn't a broken business. We're talking about hims being a business in transition right now. We're there there like there, GLP one business is completely unpredictable.

It's obviously eroding and it's transitioning that itself out. It's just at the same time of that flux, their core business is also losing momentum at the worst possible time. And just as they're trying to ramp up marketing CapEx headcounts, but they are launching some really exciting things like they are expanding into Hormona Health, which is a really big market.

I think they're layering in diagnostics with lab testing. They're launching AI Asians, like this is Ingen AI play. They're pushing into international markets like Canada and the EU and also Latin America.

I'm pretty sure they announced in Asia, like they are going global and they're not just trying to sell more pills. They're not like a pharmacy. They're trying to build a vertically integrated consumer healthcare platform.

One where a customer, uh, starts with hair loss or acne and ends up with personalized wellness membership that spans across so many different lanes like diagnostics, treatment, prevention, like it's, they're really trying to disrupt a massive industry. It's just this is their first speed bump as they're scaling up. And I think it's going to be okay.

Like you can't just go vertical. But anyway, I do want to pass it to you logical on position sizing because I think a lot of people were very over indexed on HIMSS in their portfolio because it just ran up. They did no trimming and now they're really filling this pin from this temporary bump.

How's the portfolio management going on your end? 'cause I know this is a very important earnings week for you, a lot of names for all of us. Uh, how do you like manage that?

Yeah, let me, um, real quick before I go to portfolio management, let me give a comment on the HIMMS topic. I think the elephant in the room is that they decline quarter over quarter on revenues. This is a growth stock.

You can't do that. And if you do that, you're, you get crushed. And I think something that people don't really appreciate in a business like this is, I know I'm gonna get a lot of hate for this, but it, it's kind of a commodity business, right?

I mean, they're not, they don't have any product themselves. They're just a, a distributor of a product. There could be other entrants.

People have always downplayed that Amazon risk of getting into the market. But the truth of the matter is, I mean, someone could, they have good branding, I'll give them that. Um, but I know a lot of doctors that are, you know, doing this on the side where they're, you know, building up these side businesses where, you know, you can go to them for, uh, prescriptions without having to go to your primary care doctor just to get something a little bit more conveniently.

Um, and you know, the bear case has always been, look, I think that core business revenues outside of GLP ones have actually been stagnant. And I think you're probably starting to see that in the business right now. So that's kind of where I, I'm, I'm kind of seeing this right now is revenues are declining.

A lot of the explosive growth was probably on the GLP one side, perhaps that's still going, but that revenue is at risk right now with all these ongoing lawsuits. You don't want somebody like Nova Nordisk who tried to be buddy buddy with you, you kind of stepped on their toes and now they're livid and they have the IP and the patents to go to court. Um, so I mean, you know, I I would just say that there's gonna be a lot of lobbying against you as a compounder.

You know, whether, whatever that ends up being, I, I don't know if investors are gonna sit still and watch this company, which, you know, it's, it's not that expensive given the growth, but people don't care about growth as a velocity. They care about it from an acceleration standpoint. Everyone only cares about growth businesses in terms of the second derivative.

What that means is, is growth accelerating or decelerating? And right now growth is decelerating. So that's gonna chop that multiple, especially on a business where I know they talk about, oh, this is a 76% gross margin business, but the truth of the matter is their operating expenses are extremely bloated.

The sales and marketing line item have always been high. So they've, I don't wanna say artificially, but in a sense like the gross margins are high, but the operating margins have never been that high adjusted EBITDA obviously looks good, SBCs pretty high. So that's getting adjusted back.

I don't know, it's, you know, it's one of those things. It was always gonna be musical chairs that stand at some point. And a lot of these growth stocks, you know, besides maybe a Palantir which can keep, you know, defying the impossible, uh, a lot of these end up having a day like this and I'm not sure where the stock net's out, but I would just say that, you know, potentially like some of these other growth names have something that's a little bit more durable from a growth standpoint.

Um, so that's my thoughts on hims. Let me quickly touch on portfolio management because I think it's really important. Um, you know, people love a bull market.

People love high beta stocks. People love it when their stocks go up, but the reality is that at some point it is like musical chairs or there are gonna be pauses at the very minimum. And what a pause could look like is, hey, like this stock is up 100, 200, 300% in the last one to two years.

I mean, that's beating the market on like a 10 year timeframe in, you know, a year or so. I mean, you've pulled forward a ton of those returns and you know, maybe a lot of that's been alpha, but probably a lot of it is attributable to beta, which means that there have been people who have been chasing the stock on these great results. What happens when, okay, something that people need to understand is that when you have price, it determines the market cap of a stock.

When you have a market cap of a stock, that that's what you use to determine its valuation. Whenever a stock has very high positive sentiment, people chase that stock up in price. The price means that the multiple is expanding, the valuation is getting more expensive, which means that the expectations got a lot higher.

For me personally, I understand there's a lot of people who can do really well, um, chasing, you know, what's been working. But when it gets to a point where the price is really high, then it's really tough to keep beating expectations and going up higher. So that's one thing I would say from an expect expectation standpoint, that's why I personally try to shop in areas that have somewhat lower valuations that I can feel a little bit more com comfortable in.

And I know this is a topic, a whole topic on its own, you know, shy puts valuation in the backseat, totally get it. But he does pick best of breed companies which are able to defy those odds. So you can do it, it's just a lot harder, I would say to pick true quality.

And, you know, I don't know, is this gonna be the first indication that HIMSS is really not that high quality of a stock? We'll have to see, um, in terms of position sizing and then we can move on to the next stop topic. When you're dealing with small caps or high beta stocks, things that are very volatile, something that people don't understand is that these, these small caps, high beta, they move so much five, 10% a day, they're not one 2% we're talking no news.

These things move like options on large caps. And so when you realize that, then you have to understand for your own sake of not having a heart attack every day, it's probably best to size it in a way where a five, 10% move on a daily basis doesn't, doesn't kill you. And you can end up holding through that volatility through with conviction.

So volatility is the price you pay to have high returns on the upside, especially with these volatile stocks. And guess what high upside means there could also be significant downside. And so to that, I would say, look, a lot of my positions, I hold around four, 5%.

Why is that? If I take, if I go into earnings with a four or five, let's just say a 5% position, let's keep the numbers easy, 5% position, it's down 20% on earnings, that's a 1% hit to my portfolio, I'll be fine. But when people come into these earnings and they're, you know, sizing these things at 10, 20, god forbid, 40% or whatever they're doing, and then they have potentially calls on top of that weekly ones with high iv, I mean, you're just asking for it.

And you know, so I would just say when you're dealing with high beta stocks, you're dealing with things that are, you know, more volatile in nature. The best thing to do, and I get laughed at about this because people say, oh, you're never gonna get rich with, you know, concentration, uh, without concentration you can't do diversification. The truth is that when it comes to small caps and high beta, a a small position is enough to have significant upside and a small position is also enough where if it, if it, if it implodes on you, you'll be very happy with that small size.

So I think small size is, you know, kind of underrated when it comes to these high beta volatility stocks. Yeah, I wanna, I wanna add some color to that 'cause this is a great topic, especially when the market's at all time high. So I'm sure there's some over index positions for a lot of people on, how do I maneuver this anecdotal evidence?

I'll give you all of them. So Palantir 2% position when I started under 10 bucks. What's that right now?

180 freaking nuts. I, I've don't own the same amount of shares I did under 10 bucks that I do now. I wish I did, but it's risk management.

It's been overvalued since the forties to be honest. And I think now there's validating their presence in stage two ai. But I sold half my position at 50 to 60 bucks.

I have not touched it since, even though I think it's very expensive. Why haven't I, I de-risked my position, I sold my initial risk. I'm laying the profits ride until thesis breaks.

We'll see where the market takes it. And that's what happened. Same thing with Rocket Lot.

I got over index under four bucks. I I could have again 10 XII sold half in the twenties. I and now I was laying the rest ride until thesis breaks.

So Palantir is still my top position. I've trimmed along the way. I took the profit out reallocated elsewhere, but I'm just gonna let the winners keep winning.

Same Rock Lab. But Rock Lab's an interesting one because Rock Lab has proved my thesis that they're gonna be a full end space prime, and they didn't really do that on initial pop as much. Now there's like, you're seeing their services, BI bus, there's so many hot pockets in their business where I actually kind of want to add more rock collab now, where I sold in the early twenties, I'd actually probably would love to buy those same shares back.

So that's like, uh, it didn't work that time. But either way, the way to, if you have anxiety right now and it's a multi backer, there's no shame in de-risking your original position, original capital and just letting the profits ride in compound. And I kind of, I wanna get to Palantir's earnings, but I think, Sam, any comments on position sizing before I do.

I mean, you have to practice risk management and that goes for downside and also upside, like you were saying. Uh, I mean, sterile Labs, I'm really happy that it basically tripled since the bottom last April, as we were talking about earlier. Um, and it was actually a decent sized position for me, uh, near the bottom.

So my cost basis was like around 3% of my portfolio. And then, um, when it ran, when it ran up all the way to a hundred bucks, since I was adding all the way down, uh, I trimmed probably I think like 30, 30% of my position. Um, and I've just basically let the breast ride.

And I've considered trimming over here, but it is just, you know, at this point, around three to 4% of my portfolio, I feel like I'm just gonna leave it. And, you know, the, the, the research that you guys come out with in future equities, totally agree with all of it. It, it just seems like the, now I don't, I don't wanna say we're getting started on the narrative per se, but it just seems like since we're very early when it comes to ai, in my opinion, and when you think about the bottlenecks that come in with network throughput, with a lot of inference queries that are running, I, I don't, I can't last time, like trading around the Yeah, go ahead.

For the audience network throughput, what's your interpretation for that? For me, it's this data movement between memory and it's just latency speed. But describe to everyone else in a well that, that Would be LA speed.

So you have, you have the homes, I would say, as far as individual servers in an environment, and then you have the data that runs through them, right through the servers, whether it's network cables, whether it's timers or whatever it is that Sterile Lab has. And that's where their mode is, right? They, they ha they have the components that are used to build these GPUs and GPUs clusters, but they also have the timers that are used to basically increase the speed or redo the re-accelerate the, the speed that these data packets are running through the system.

So think of that as the highway, right? You have the highway, which is the wires, the physical wires, or which is the wireless signals that run between each one of these, uh, units. But then you need these retirements in order to reaccelerate or give it that little boost to be able to make it as fast as possible to the other end or to the host, to the destination as soon as possible.

And that's what happens with inference is that someone's sitting on their phone and they, let's say they're using chat GBT, they ask chat GPA question where the information already exists in their database or in their LLM, but they kind of need to work with the existing data that you do have on the already trained model. So instead of training, we're thinking inference. And that information needs to go back and forth between chat gpt as well as the phone itself, as well as other parts of the database that needs to hit.

That is inference. That's how inference works. You're not teaching the model how to speak, you're asking the model to translate something for you.

And that is what matters in all this. Because like you, oh, go ahead. Go ahead.

Sorry. No, I was gonna say, so add more color AI's intelligence. We can all agree AI just is intelligence.

That's why it feels like there's going to be an indefinite demand for all all this AI thing because it's not a product launch, it's a characteristic change on the society who knows the ceiling on that inference is, is the tax on ai? It's as simple as that. So like what Sam was saying, tragedy inquiries or every AI workload, it's inference.

The tax, the tax you have to pay to use ai. That's why Broadcom is the probably the most quiet trillion dollar company in the world that nobody realizes. 'cause they're Uncle Sam of ai.

They are, they collect the tax on everyone's AI workload through networking, working es stair labs, es stair labs essentially is, I'm trying to think of a better term. I the toll, um, the toll booth for example, like, uh, what's that phrase in the California highways where you go on a toll? It's, it's, it's freak it anyway, it's a toll booth.

You got, you go on a highway, you have to pay, we get you. So no, I'm saying for, for the audience A but a highway. Yeah, no, there's like a specific fast section on the highway that you can pay.

It's like a lane toll road. It's, I guess it's just Road express lane or toll road, whatever it might be. Broadcom essentially allows every single car on that highway to go through their toll booth.

They'll collect tax on every single one. ER labs is hyper specific to the spec AI bottleneck that's happening right now. The Ferraris uh, out there that just will go on that one lane and just go from point A point B super quickly.

That's why ER Labs is kind of a derivative play on this networking theme where Broadcom is the king, absolute king Aire Labs is just the purest form of capitalizing on what's making Broadcom do so well right now. And the only issue that I'm curious on your input on this, Sam, is we all know what happened in Confluent last week. Maybe we didn't talk about it, we won't get into it.

But the AI tsunami that I constantly talk about, it happened to confluent last week, the first glimpses of it where their biggest cu one of the biggest customer, maybe the biggest, I don't know, open AI, drop them. Essentially they're gonna use their open source to create their own in-house competing product for, uh, streaming data. And they're like, why should we pay for you when you have an incredible open source product?

We'll create our own AI agent solution to compete with your managed services and we'll, we'll be done with you. I feel like Sarah Labs is somewhat of a risk for Nvidia to stop being buddies, buddies with the stair lab if they get big enough. And that's always the worry.

'cause they're so reliant on Nvidia right now that if a stair lab becomes like a 50 billion, a hundred billion company, you never know when Nvidia will just create their own competing product and cut them off on any kind of envy linked derivative. Like do you think there's that that's a risk at all for Sarah Labs or am I just being too paranoid? I feel like it's very different.

I mean obviously when we Get to valuations that Ster Labs is trading at, you know, it can become very sensitive to any type of negative news that comes out. Since everything is just good news, it's all balloons, parties everywhere. For Esther Labs, every Wall Street analysts probably very excited and hiking the price targets as they always do for Wall Street.

'cause they always chase up and they always chase down sort of. Um, so yeah, I mean I wouldn't be surprised that the Sterile Labs comes back and visits like the a hundred dollars mark where it found a lot of resistance before. But from a fundamental perspective, confluence is software.

It's the stickiness of software products is is a lot more impact as far as budgeting goes, as far as headcount goes and scale-wise then changing your hardware, right? Especially a critical component that Sterile Labs does offer. In order to change that out, you'd either need to build it in-house or you'd either need to outsource it to a different company, right?

So let's just say that there is no other company that does it and they're likely is not another company that does it as good as Sterile Labs as we know of right now. Nvidia would have to do it in-house. Now if Nvidia were to have to in-house, they're talking about architecting and designing a complete different component that they used to outsource, I think we would probably have wind of that ahead of time.

And if that were to happen ahead of time, I think Sterile Labs probably knows that as far as management goes, that they're trying to diversify where their income flows and clients are, which they have already been doing. But also on top of that, like I was saying, was that switching from Confluent to building your own in-house platform as far as data streaming goes is something very easy to do for a lot of companies, especially when you have billion dollar engineers that, uh, Sam Alman has said that they didn't go to Meta, they ended up staying there and so on. But also, uh, a lot of companies like Meta as well, they don't really buy too many vanilla products or off the shelf products.

They build their own concept. They, they basically fork a branch of whatever they're trying to do and they just build it in-house. It's hard distributing in-house.

They save a lot of costs in doing that. To do that with software, especially with Confluent, which Jay Krebs who's one of the founders of Kafka itself, it could be easy to do that for them. And they probably were working on it for years and they finally figured out a way, Hey, we don't need to use Confluent anymore.

So they would just drop them for that. This could be happening. It's actually very interesting because we were talking about this before with Datadog, right?

Datadog could be in the same situation. Datadog is not an open source platform, but there are a multitude amount of software platforms in terms of monitoring, security, whatever it is, including Prometheus as far as visualizing graphics, also Grafana as well. Grafana is open source and a lot of these hyperscalers offer their own platforms as far as CloudWatch for AWS in fact, I would even say firsthand, not, not just them though, cybersecurity too law of cyber security platforms have observability.

So it's easy to switch off of software. It takes a little bit of effort, takes a little budget switching, maybe recode some of your applications for the dependencies and everything, whatever. But to switch a component, a critical component in GPUs that you're trying to meet demand with, with building a lot of supply ahead of time and still being short the supply, that would be a very ballsy move for N Vdu.

Even Jensen Wong, maybe within NVLink. That was something possible. That's something that they've always been working on that came up with env ENV Link Fusion.

So then that way people will stay within their ecosystem, that GPU clusters by offering compatibility for other GPUs offered by other companies, including TPUs and so on. But to do that with Ster Labs, Anything they partner, they, they partnered with Stair Labs for that. So like they, they know how valuable Stair Labs Scorpio is that they're just gonna partner up with 'em, which is a great sample of approval.

But that was announced by the way, stair Labs was, is in the sixties I think, or seventies crazy Opportunity. I it didn't really react like it didn't really react like it that that that was, that was some time ago. And also like, you know, they went below their IPO price, which a lot of people were saying like, oh, if I bought an IPO whatever and people had the opportunity, they didn't take it right?

So, you know, I'm not saying that's a the only missed opportunity on the market, but when we think about just the opportunity that's been presented three months ago versus today, there's not a lot of opportunity today. I mean I obviously there's a lot of, there's a lot of smaller caps and mid cap stocks that do have a lot of opportunity, which Logical has been taking advantage of a lot lately and has been seeing a lot of upside on. Uh, but as far as like mega caps, like well-known brands that a lot of people know, like a lot of those opportunities are really just gone.

Well You, you, you brought up a great point. What's the next opportunity? And I think Logical and I both share one name on where that opportunity is and it's going through a rereading and this, this week's earnings prove that I'm gonna let Logical take it and I'll add on, finish off my thesis on them.

Uh, hold on. It's, are We, are we talking about Digital Ocean? Yeah, we are.

We're swimming in digital Ocean. I, I'm, I'm sorry, but we, I already told you I cut it today. I I know, I know, I know.

But you're, you got, you got the post earnings gap, you had the set up, but you got the 30% move. What, what attracted to you? Yeah, What attracted me to it?

Um, it's always been an interesting one in the compute space. It, it just kind of operates where the other ones don't. Um, and the valuation always looks so good.

I mean, for me, look, I I've been having a little bit more of my trader hat on and I noticed that, you know, this thing is basically trading near its lows of the range. Uh, valuation looked very reasonable. The reason why I was alerted of the stock is because I saw a lot of, uh, options flow bullish, options flow come in and I revisited the name and it made a ton of sense.

And so I just took a position and it's a name that I always wanted to own and we've talked about a lot, but it was back to a point where it made a lot of sense ahead of earnings and it worked out really well. But honestly, you're the expert on the name, so I'll let you talk about it. Okay, yeah, I'll add some color and how dare you portray me.

Uh, we might have a new co-host guys next week, honestly. How, how dare You believe me, it was a mistake. It was a mistake to cut it.

I'll say that right now. It is going higher. I can feel it in my bones and it was just portfolio management strictly not anything against through motion.

Yeah. All all good. Yeah, I mean I've been in digital OSHA for far longer than I want to admit.

It's definitely been not the be best stock to own. But I did double my position in Digital Ocean the past couple months. 'cause like what Sam was saying, the winners, the or the proven winners have gone through that multiple expansion opportunity already.

So it's becoming more and more expensive to own the no-brainers. So if you have capital coming in every single month, you, it's not wise to continue adding to the proven winners at this current multiple. You have to look elsewhere.

What's the next opportunity? Cloud computing was on fire this cycle. You got it from Azure, you got it from Google Cloud.

AWS is still really, really, uh, doing well. Like they exceed expectations. They had a, the head of AWS completely messed up a couple years ago, maybe it was a year or two ago.

He went way too hard on train him and he got fired for doing that. And now they're just compute and energy constraints. And that's why they kind of didn't exceed expectations as much as the Azures and Google Clouds did because they uh, they couldn't fulfill the demand, like their backlog's growing at 25%, their top line's at 17 or 18%.

That's a huge fumble on management and I think that's why that person was let go. But it's not a demand issue. I don't wanna call that out.

But either way, all three, the demand exceeded expectations. So there is a hot pocket right now on cloud computing. DigitalOcean is a name that I always had believed would benefit from the explosion of SMBs that are gonna see from AI agents, uh, creating these companies overnight.

I can have five SMBs in a matter of like a week just having an incredible AI agent create something for me and I need it. Don't want to pay up the prices for the big three, I'm gonna go to Digital Ocean. But obviously business spending has been ta uh, put a bit anchored and a small business space due to high tariffs, um, high interest rate environment.

But either way, I think that I saw digital Ocean trading at 10 times ebitda multiple management has called out a couple times that they expect their top line growth to end 2027 at 20% with while maintaining 35% EBITDA margins. So either the street which is guiding only 14% top line growth does not believe what Patty's saying to the CEO or they're just like looking, they're overlooking this AI cloud that nobody's ever heard about. And that's where the opportunities are always found when there's a massive disconnect from what the market's saying, this is what you're worth and how many eyeballs are on it.

I love just like risking the biscuit on those kind of names. And this is the cycle where I think they're going to go through that inflection rerating, they beat it across the board. It was a one of the cleanest piece they've had in years.

I've covered the stock for years, but the real signal wasn't the top line B it was the mechanics that were underneath it. They had the highest a RR in three years this past quarter that tells you everything you need to know about the momentum that's building in their cloud computing offering. It wasn't just from one segment either.

It was both the core cloud. Uh, and it was also the AI that contributed me meaningful to that number where the AI machine learning revenue by itself doubled a hundred percent year over year. That shows that this, there is an AI story that's being had right now.

And also I think it destroys the bear case that this is just a one legged growth story. It's not AI is providing a renaissance for their offering. And I do think that there second Bear case is net dollar retention sucks for the same, it's awful.

It's never been, it's been a while since it's been over a hundred percent. I think it went from 97% to 99%. And I think it's really important to know that this net dollar retention is excluding the AI revenue.

So lemme say that again. AI is not included in their NDR calculation yet. They're still improving cohort behavior and raising their forward guidance that tells you everything you need to know that this is getting, going through a re-rating, it went up 30% after their earnings.

What did it do the following day went up 6%. The follow through days are vital to see if a move's actually valid. And if there's a re-rating that's happening, especially on a 30% move on a name that just isn't highly shorted, it's just left for dead and nobody really cares about it.

Shows it's got the, it's got the analysts attention. So I think that, um, DigitalOcean is one of those names where if they stick to their guidance in 2027 and they continue this momentum, I think there's gonna be a massive rerating towards like 25 to 30 times ebitda. And I think there are EBITDA right now.

Let me pull it up real quick for 2027 is, wow, it is 600, $500 million. So I think, wow, this is going to be a 15 to 20 million company if I assign a 20, 20 times EBITDA multiple. What's the train now, right now, after this move 3 billion.

I I think this could be a five Xer in the next couple years. I really do believe so. And just because it's just so severely undervalued.

AI is not just a product cycle bump, it's an indefinite behavior change. And I just, I think that that alone combined with the fact that I think they said somewhere between, I think they said 10% of their business now is AI machine learn learning revenue correlated and it's growing at a hundred percent clip. Wow.

And I do believe that they're building this full stack agentic AI cloud now as well because that AI business is really picking up steam growing fast. And the customers are buying in. They, they're investing in it.

GRE AI is live in GE in Atlanta, Georgia. I think they're AI and um, they're, I think it has like already 14 or 15,000 agents that are created by over 6,000 customers already. Ha Like live in Gradient ai that's nearly a third of whom are like net new.

I'm pretty sure just to digital Ocean. Like it's accelerating. It's not just an add-on product.

I think it's a full new customer acquisition challenge channel. And I think it's directly tied to the info base. And I also want to call out, they sold 76 customers from Hyperscalers last quarter.

It's not just for SMBs. Like there is IT movement right now on inference when we talk about, I know I'm like, uh, dragging this along a little, but like inference is essentially letting the small guys compete with the big dogs now. And I think that you're going to see that with a MD they're catching a win from that.

You're going to see it from a lot of these like companies who want the cheaper offering that has a similar-ish product. 'cause inference is all quantity. It's a tax.

You don't have to have the highest quality on paying your, uh, paying the tax on AI workloads. You don't. It's just like an add-on.

So you go for the cheapest offering. That's why I think this is a massive tailwind for DigitalOcean. And uh, right now you're seeing the price act accordingly.

So, uh, I don't know, what did I convince, uh, you to go back into a logical Sam? Are you, Sam you're in the tech space? Like Yeah, what you just said.

I'm like looking at the chart and I'm like, I think I blundered. It's, it's a $50 stock. I think it's a $50 stock this year.

It's a like next Couple months. It's, I think, I think what I would push back on is like a hundred percent growth sounds great on that AI side, but if it's only 10% of revenues right now, I mean you still need a lot more growth to make it more meaningful. Right?

So it's still gonna be like at a hundred percent, you know what I mean? I don't know. I'd, I'd, I'd wanna see it.

I mean if they conti continue to go at that clip, then yeah, clearly that's gonna be huge. So I mean, my thoughts on Digital Ocean and I, maybe I'm just biased because I I've held a lot of like smaller mid-cap companies in the portfolio and of course larger cap or mid-cap, large lar lar, larger size mid cap companies and a lot of leaders as well. And I'm just so biased to owning the leader over anything.

Um, you know, when I think of like small SMB uh, companies that focus more on SMB versus the leaders, you know, I think of like CrowdStrike and Satel one. That's the first one that comes to mind. 'cause I've owned both.

Um, obviously Crowd Check has seen a lot more upside than Sentel one. In fact, I would say that in the last two years, crowd trek's up basically like 300% while Sentel one is up like 10% max and has massively underperformed the market. That's 'cause seven one really focuses on smaller businesses and also smaller businesses can be very volatile in terms of their, uh, clientele in terms of revenue that they do project and everything.

It gets much harder for them to guide. I'm not gonna discount digital lotion. Uh, I mean I obviously need to look more into this.

I've always, whenever it comes to hyperscalers, I'm always more leaning toward owning the larger sets versus the smaller ones. Um, But it's such a, it's such a big pie though. That's the thing.

Like even if a sub, Its a 5%, A big sub, 5% Is massive. A big Yeah, it's like nebulous versus uh, AWS or something, right? Like it's such a big pie to eat.

Mm-hmm. And the tam is growing very quickly. So the smaller place can, can make a lot of money.

The risk reward is obviously the upside of digital lotion. And I'm not saying that it wouldn't buy right now, like I'd have to do my due diligence to look into it. I've owned this in the past, probably like earlier this year and traded it per se, uh, got out of the right time, um, was considering getting back in, in the mid twenties as a trade.

But I, I've never pulled the trigger as far as the front about the basis. But this is definitely something you need to look into and I really have no problem buying something after 30% do It, do it, do it, do it tomorrow. It's like, it's right now it's Likem You into it.

It's Like tine, like do it. Um, But Hey, hey Pat, pat, Patty, if you're listening to this, we love you to have you on. I know you follow me.

Yeah. I should probably DM you like that would awesome. Im a big believer.

Yeah, we should we still swim in the pond that you're working on right now? Yeah, that'd be really cool. I mean, you're really coming head to head with a lot of the hyperscalers out there and I don't know if it's gonna happen, but it could be open acquisition at this point, like three, $4 billion company.

And then you especially have these larger companies that big valuations that could potentially be the, uh, inquisitor, whatever the hell it's called. Like who's, what's the, what's the, what's the name of the person that acquires the acquirer? The acquirer?

Yeah. Um, we are talking about this before and we were speculating on it. I mean, what if it would make sense if CloudFlare did?

Okay, I'm sorry, I'm looking at a chart right now. And Apple is just ripping to like the two twenties right now based on the news that just happened recently with, uh, Trump basically saying Apple is exempt from the, uh, semiconductor tariffs and so on. That's Price in though there up 5% running after hours.

It's up 3% after hours, which brings it up to 3 billion trillion. I thought there was another piece of news They're getting new Manufacturer. GL what Is it?

Oh, you're talking about Corning, gw, yeah. Yeah. Corning.

Corning, yeah. They, they, they're, they're, uh, expanding their manufacturing create and not creating four to 50,000, but they basically support that many, no, they, they're getting a lot of positive news, but also you have like these laggards in the co in the market that, and I was just talking about it earlier, um, that are gonna make a comeback, right? And when you think of Apple, it's like, that's obviously a very well known brand and it's gonna make a comeback as far as a narrative goes there.

I don't think there's any reason why that Apple should not be going up to the market. But anyways, um, going back to digital lotion, like I, I definitely think they can make a comeback when it comes to that. They don't have to worry about open source software or anything.

And I don't know, who knows? It would make sense if CloudFlare acquired them. I don't see why that wouldn't happen, especially if CloudFlare did sell some or raise some money, you know, by, uh, diluting some shareholders recently at, at their crazy valuation right now.

I mean, what's $4 billion of a $74 billion company? Probably more than that today. Like 76 billion.

Like it could happen. I could see Matt Prince doing it. Like it makes sense, right?

If you wanna scale out your CDN, which is probably one of the most important things when it comes to the inference world. Why not acquire all this free compute that you could use yourself Millions of developers they to get as well. But here's the caveat.

I do wanna get to the next topic, but add color. I don't think they're gonna sell out because they wanna AI talent spree hiring spree a year or so ago before they had a pay up for it, the, their C-T-O-C-P-O, his previous role, they poached him. He ran a, the AI division from AWS before that VP of Nvidia.

They, and I forgot who they hired after that person. Like they, he Patty just wanted a hiring spring. There's talent around.

I don't think someone like that leaves running AI for AWS to work summer for two years and get sold. And maybe I'm a little naive, but either way, let's talk about a next hot topic. SAS is dead.

And I think that's a fun one because I think SaaS is very dead. I talk, we talked to, um, bill McDermott that serves now two weeks ago. We've talked to a couple people in that space where being a SaaS company is now having the nicest house in the worst neighborhood.

Nobody wants to know that you live in the SaaS zip code because that AI tsunami I'm talking about, it's coming for that whole space. I don't think people realize 75% of the SaaS companies will not be here publicly traded in 10 years. I think they're gonna go, it's gonna be a heavy cannibalization.

I think we saw signs of that with Confluent, but I think it's just more than a confluent issue. I think we're watching the SAT stream confront a quiet reality where great products don't always translate to great businesses, especially in the, in a world where we're in right now, the strongest products are open source and the smartest customers are AI needed open AI confluent, we just talked about it, but it's not just a confluent issue. Elastic, they have an incredible search products, but their managed services is disappointing.

Like everyone uses their open source search product, but they're growing the low teens. But let's talk about confluent for, for ex uh, first because they just had their last, uh, their earnings last week. They're trading at $17.

They're just getting destroyed. They have an incredible management team as well. They have a phenomenal product.

I think we can all agree that streaming data's very important in the new digital economy. And Kafka is so good and so embedded and so widely understood, uh, that Fortune five hundreds, especially the AI native ones are saying that why should we pay for your managed services when we already have the talent in-house to run this ourselves by using Kafka? And OpenAI made that conclusion, uh, already, I think there's gonna be more bigger clients who are gonna follow suits.

And that single customer OpenAI forced a down revision to the back half of the year and undercut the entire cloud native consumption narrative. And now, although Confluent is this still this mission critical still foundational product via Kafka, it's in the hands of these elite technical teams and it's is increasingly being treated as infrastructure to own and not rent. And I think this is really important to understand the difference.

I'm curious if you guys have any thoughts. I I just wanted to, can you just like explain a little bit more about data streaming? 'cause Yeah, So here, do you have the technical uh, answer to that Sam?

I'm gonna bring up a DOMA five version. Okay. So of what it's, When you think of a database, it's data at rest, right?

The tables and whatever the values are. And every key, key value, if you think of an unstructured database, it sits in a table on a database, right? And it just sits there and DA and applications hit these tables and they process them and everything.

But the data is just sitting in the database, right? It, it might be processed through an application or script, whatever it is, but it's being put back into a static data database. What Confluence does, what con, sorry, what Confluent does is that it's taking data that's in flight, so data that you're sending from point A to point B, but it's processing that data while be, before it reaches its destination.

So if you send over, um, if you send over two plus two over to a destination, confluent will say equals four and then send four over to the destination, right? In a very basic sense, that's what confluent, that's what COP does Communication. It does the communication layer of the data.

It's, It's more of a, it got really popular because of the live, uh, ride tracking and like the credit card fraud detection message Streaming as well. Like All, all the stream like that. It became like a niche that obviously grew because it became so reliable and so well known by the big companies that everyone just essentially used it.

So how Are they able to phase them out then if it's like so important, like how is this such a easy product to replicate? So because they have an open source product, right? So if you think of, uh, let's say databases as well, um, let's say MongoDB, right?

MongoDB is open source. A company can either pay MongoDB the company, um, hey, I wanna use your product and I don't wanna host your infrastructure. I wanna, I don't wanna host my own infrastructure or databases and maintain them.

I wanna pay you and I'm just gonna use your service. That is SaaS. That is software as a service.

You're servicing software not in-house. You're just paying for the software to manage it externally, right? It's kind of like your email.

You use Gmail, you're not hosting a database server, you're not hosting an email server locally. You're not hosting the endpoints and creating APIs doing it. You're literally paying Gmail to hold your email for you and just use Gmail for sending email, whatever it is.

Confluent is the same thing where they're trying to push Confluent cloud, which is basically they're managing the infrastructure updates, upgrades, maintenance, whatever it is for you. You're just paying them for their service. Kafka Open source is something that you take like an executable or zip file.

You're installing an application or whatever, or Kafka locally. You're managing all the infrastructure, managing all the processes, whatever it is you upgrades and everything locally. You're literally spending your budget to pay people to manage this for you and to be the experts in configuring the entire application integrations, whatever that is.

You're paying your own engineers to do it or rather you could pay Confluent to do it for you. That's where the benefit is. The problem is that you have big conglomerates like OpenAI where they have a billion dollar engineers who could figure all this out in their own by just forking a Kafka project for pennies on the dollar.

It's, it's essentially that everyone can, anyone and everyone can create their own lightweight versions of what Confluence managed conf, confluent cloud is. And because Kafka's such an open source product that everyone can use, that what I'm referencing, like these bigger whales, they can create 70% of what this really great expensive product can do. That 70% is usually good enough for them.

And I think that's I, okay, so I just, 'cause we're on the topic of SaaS and I know that like Sam and you kind of have different opinions on a lot of these names. 'cause I'm trying to get to the meat of the conversation, which is like, SaaS is dead. That's, you know, your take, uh, Sam, uh, shy.

But you know, Sam, I don't think you think that I'm kind of in the middle. I'm not sure what I believe. I feel that there's still gonna be some SaaS and maybe not others.

But I, I've always thought that software is becoming a commodity. Um, I think it just becomes later on about the integrations. But seeing something like a confluent get, you know, disintermediated, it does make me, you know, go a little bit on shy side the more I think about it.

I mean, what are your thoughts? Because I know you like GitLab for example. I mean, what are your thoughts there?

Okay, so GitLab is also open source and that's the issue here, right? We're talking about GitLab cloud or they, they just call it GitLab SaaS where they're hosting your entire infrastructure on their services. And specifically they use AWS for most of their clients and then they offer it to people, right?

That's their driving product. That's what all these companies are making higher margin on with their scalability is to host it themselves. And since they're hosting themselves, they can pay AWS or Azure, whatever it is.

Like, hey, we have like 500 clients who are hosting stuff with like give us a discount, right? But if you host it yourself, you're just paying for a license. So there isn't that recurring usage cost that I know shy is more bullish on for consumption based companies versus license based or SaaS based.

So the issue here with GitLab is that, well what if people choose either GitHub, which is Microsoft offering for uh, DevOps platform or what if they just fork the project and just build their own right? Which a lot of people have done. And also there's many other open source products out there and there's many other competitive co competitive products out there.

My argument with GitLab per se is that if you think of a pure play in terms of DevOps, your choices in an enterprise environment are really just GitHub or GitLab, right? Tho those are the two leading ones. Microsoft is very hard to compete with 'cause you had the bundling power.

But also on top of that, GitLab's metrics really show that they're doing really good other than the fact that their net retention rates did not, not the net retention rate, it was their, uh, clients under, um, $500,000 or 5,000. I forget the exact number, did not grow as quickly as the market expected and therefore they sold off in the last earnings. Very volatile stock.

Smaller cap as well under $10 billion. But at the same time, yes, there is a bear case about it. I understand that and I am battling that with risk management with the position size.

But go ahead, shy. No, I mean you did a great job. I think, I just think for SaaS being dead, you have to ask yourself what's middleware and what's a actual infrastructure, data infrastructure.

And that's a really important difference where middleware is gonna get replaced. I've already heard from like the, like bigger fish, like they're, these bigger companies are becoming more and more wanting to control everything in their stat vertical integration. Like they wanna control everything in it.

There is an opportunity now that AI is gonna be able to offer them that they can explore creating these competing products. They weren't able to before 'cause of cost efficiency issues or, uh, they didn't have enough headcount to do so. But now these AI agents could help through that process.

And I think that you need to own the data now or in some sort, like have data in your ecosystem that's protected in order to really survive like MongoDB. I, I, I think MongoDB is, uh, a name that I just think is evaluation play. I don't know if they're gonna really do well.

'cause that open source theory ahead snowflake, 40% of the Fortune two thou 2000 data is a ReSTOR in the ecosystem. That's a huge lever. But that's sa uh, that's SAS's that conversation.

We can probably, we circle back to that, uh, throughout the weeks, months. It's not just a one-off. This Conversation is dead.

Yeah, we do want to talk about Amm d though before we do start. Yeah, real quick. Sorry.

Sorry. Just one one more thing real quick. Um, just because you know, confluence issue has been the OpenAI there was their customer, right?

That's and they dropped them, right? But does GitLab have a similar risk here of that They don't have a similar risk that They don't have concentration client risk? They, They, they don't.

So that's Honest big deal, right? I I mean I think that changes the calculus a bit. 'cause 'cause that's what the issue was with Confluent was they had, Confluent has always had that issue.

They've had a, so there back in 2023 with where they had two clients that one of them was dropping them, the other was deciding getting off that was 40% of the revenue that was the issue back then. So that posed the risk here and it was not good. Well, GI Lab just has a GitHub issue.

That's the primary thing. And I think Microsoft's gonna own a lot of the pie. It's fine.

It's it's $7 billion company. Anyways, I I do wanna add though, confluent is still trading at 40 times EBITDA multiple after this drop. So you're saying short it No, I'm just kidding.

No, I I'm just saying like one customer flip a switch off and they're still not cheap even after going out 45. Well, that It's, well, yeah, but that's a big part of their volume of revenue. That's how you get negative operating leverage at that point.

The valuation actually goes up in that sense. Right? Exactly.

That's why, That's why I'm bullish GitLab because they're not really in the same boat when it comes to that, right? Like obviously GitLab is not profitable on a gap basis, but like, dude, their metrics look great. Like it doesn't go ahead.

Here's the, here's the thing about GitLab though. Like, everyone knows that AI tsunami is coming. It's the world's worst kept secret.

They're gonna get acquired by Google. They're not gonna survive in five, 10 years by itself. That's great.

Yes, please, no, but they're, they're never gonna get a premium multiple that they deserve because of that. You're seeing like the proven winners, stage two AI winners, CloudFlare Axon, Palantir, ServiceNow, they're all maintaining their premium multiple. But the rest, there's a huge distance between tier one and tier two soft brands on what the multiple should be.

There's a reason for that and I think it's going to continue widening and it's from the AI tsunami life. But, but again, that's enough SaaS 'cause we gotta talk a MD right now and I think that lo logical made a really interesting play. That was AI a MD adjacent.

So talk about a IP before we get into a MD. Yeah, so A IP is our terrace. They're basically like a chip design company.

Uh, and then they had this basically like a PR announcement where they're working with A-M-D-A-M-D has selected them, um, as basically a partner, uh, to design a lot of their chips that are gonna enable ai. Uh, basically they have like these, uh, network on chips things, which helps their infrastructure in terms of, you know, being faster computing, more capable, et cetera. Um, the stock popped 50% after hours, which was interesting.

And that was getting a lot of like, uh, volume and the stock's been looking good, and then they had their earnings report and the stock gave back all of those gains. I'm glad I took some profits, uh, on a trim, but then I listened to the earnings call and they basically said, yeah, so people were like, you know, prodding and asking, Hey, what's going on with, you know, the a MD revenue? What should we expect from guidance standpoint?

And they're like, oh no, that's already been baked into guidance previously. So it's like, wait, what? So There, That's why the stock did not hold any of those gains because while they have like this fluffy PR um, it's not really showing up in the revenue quite yet.

Um, but it's, it's still validating the thesis. So I still like the stock I added more back, um, when it came back to reality. Uh, it still shows that it has a very bright future and it can lend more of these contracts.

It's more like partners end up like vetting their technology and stuff. So anyways, I'll pass it to you guys, but that was the small news on the A-M-D-A-I-P front Is Sam, is your, is your A MDA positive or negative take? I think that negative should go first.

No, I I am, I'm bullish MD I I just don't own the stock. Um, No, no, I meant, I meant earnings reaction. Oh, I mean, what what's your take from the earnings?

You know, there are a lot of expectations when a, when a stock essentially doubles, you better show up, right? Um, data center didn't do as well as the Shri had hoped and it got hit. You know, when, if you take a lab for example, or stair labs, if they didn't hit on their analyst estimates, they'll get hit.

Even if a lab was in line, they would still get hit because they came into the earnings with a crazy, well not crazy valuation, but with an elevated expectation of what the company was gonna perform. People thought Lisa Sue was gonna come back hard, this earnings not saying that she won't. I think a MD is gonna be an amazing play for the next five years.

Uh, that day just was not today. They did attribute a lot of the reduction in growth for data center to the chip restrictions, but at the same time, that literally was just lifted a month ago, right? Which didn't necessarily include most of the quarter at all.

Also, on top of that, there was a bit of front loading in the first quarter, and then you came into the second quarter, which Ernis was part of. You saw that hit with Nvidia. I kind of expected the same with a MD, but I didn't expect the company to be down 11% today.

I don't know exactly what it was down today, so don't quote me on that, but it was down pretty big today. Um, I think it deserved a little bit of a shakeout from a technical perspective, but also on top of that, a little bit of reality. Like, hey, you know, this is not the leader in the biggest wave we've ever seen in technology in our entire lifetime, so let's treat it like one.

Right? You saw Nvidia do the complete opposite. NVIDIA's up.

I think that was basically the affirmation that Jensen Wong was like, Hey, you, you know that, uh, you know that meme from um, captain Phillips where he is like, look at me, look at me. I'm the captain now Captain. Well, basically Jet Wong said that, Lisa Sue.

But instead of saying, now I'm still the captain, all right, I feel pretty bad for a lot of people thinking that a MD was gonna pull in Nvidia. It's not going to, I'm sorry, but let's just say gaming comes back and PC comes back, all right? A MD will see a massive tailwind if that happens.

But until then, NVIDIA's still the leader. It is back. That's the thing.

I think that was the, that was the biggest hidden gem, not hidden, but it was the biggest gem that earnings report was CPUs and gaming. GPUs are carrying the story right now, but the issue was like what Sam was me mentioning. It doubled.

So the valuation was screaming A-I-G-P-U supercycle. It wasn't that narrative, this cycle. It really wasn't.

And we talked to Lisa actually before the earnings and got announced and we communicated that she had to be, they have to be overly communicative on what the inference tail one will be on their a i data center space. They can't just call out the A IDC. They have to specifically say what inference is doing in that section and provide more layers of all everything that's happening under the hood on the data center front from ai.

And I think that what's happening right now is justified for the stock price just because it's kind of got ahead of itself. It happens all the time. Like everyone thinks that a MD is going to be closing the delta with Nvidia, uh, but it's not trying to be the next Nvidia.

It's trying to get that sub 10% share of the AI accelerator market, which is a huge opportunity. And to AMD's credit, their ex execution has been world class. Lisa is probably one of the best non founder led CEOs out there, especially in something as important as what a MD does in that semiconductor space.

It's one of the few companies that actually mastered chip based design at scale, intelligent really do that. And I think that's, there is a world where a M D'S TSM partnership is gonna be at a level where it's essentially becomes a co-development thing. Their roadmap aligns inference really does scale up.

The China heartburn is gonna get digested, it's gonna move on. But for me, like I wanna see more from instinct. I wanna see Oracle type wins with hyperscaler size, uh, numbers.

I want visibility in the MI three 50 adoption beyond what they're saying on the conference call. If a MD has the second best inference platform in the world, which I believe that they do, then that $50 billion TAM by 2030 is up for grabs and stock might look a bid, uh, overvalued right now. But if AMD starts to prove again, like they get that sub 10% share of that AI accelerator market, I think it's gonna grow into that valuation some more.

It's just, it feels like they kicked the can down the road for Q3. Like the burden of proof now shifts to Q3. And I think the stock just ran up way ahead of its like what's was actually gonna get communicated and delivered.

It's nothing Lisa's fault. Stock moving is all based on retail, all these different dynamics. That's not whatever she was saying.

So, uh, yeah, It's not, it's not her fault at all. I think Lisa Sue's one of the best CEOs on the planet. Um, yeah, a lot of people laugh at her because of the stock price, which is just stupid.

You don't blame A CEO for the way the market looks at your company. You, you, you determine A CEO based on the fundamentals of the company over years. Not a quarter, not two quarters, not even two years or three years years.

And she's been at the helm for a while. She turned the company. Remember when, if they really wanna talk about stock price.

Do you remember when a MD was in the twenties? That, what was it like five, seven years ago? It was in the twenties and people thought a MD was over.

Look where it's today, it's eight times more that price even after this pullback, right? She's a great CEO. She deserves CEO of the year, uh, what was it last year?

2024. But, um, before we continue, actually, I wanted to step a bit into, uh, the medical AI space. Um, we have on Friday, Tempus AI's reporting, uh, the Nancy Pelosi bet, uh, which is actually pretty interesting because, um, I'm a holder of Tim.

Uh, not as big a position as I would want it to be. I want to get the opportunity to add to it, but there's also other small players in there like PSNL, uh, logical. Do you wanna talk about that one?

Yeah. And just to close that, a MD convo, two quick comments. One, I still can't believe Lisa, Sue and Jensen Wong are cousins.

That blows my mind. Um, and then two, uh, you know, while it was like a big dump today on a MD dude, I just looked at the chart, it's beautiful. I mean, it, it is, it held the 21 EMAA wick below it and held it and it's huge volume.

So people bought the dip today. Anyways, um, moving on to PSNL Personalis, um, Natera, who is another, uh, testing company and then Tempus ai. So personnel reported yesterday, Natera reports tomorrow and Tempus reports on Friday.

These are probably the biggest, uh, players in like the AI healthcare detection test space, whatever. 5% or something. I never go too big on ideas that still have some binary risk.

They missed the mark, they lowered guidance. Um, they basically talked about how there was a decline in revenue for pharma tests and services. So I feel like that's gonna be a read through going into Natera tomorrow and Tempus AI on Friday.

Tempus AI stock has been extremely weak. Short interest has been ticking up even as the stock has been struggling. Um, and then Natera reports tomorrow and you know, personnel has some partnership revenues with them as well.

They partnership revenues with Tempus AI as well. So it's kind of like, I'm wondering if this is a readthrough. I almost wanna short these stocks seeing what happened to Personalis.

Uh, I'm, I feel like some of these reports are gonna be out by the time I, you know, you hear this, but I mean, I'm just talking man, I got absolutely crushed. Personalis is a small cap. It, it requires insurance coverage to see those volumes and to be able to see that margin expansion opportunity.

Uh, management is kind of, I don't know, like they talked it up really confidently in the last call, this call. They're still confident. They're like, we've submitted three things now.

We expect two, you know, we have three shots on goal. We expect to get two coverages by the end of the year for two different indications. So, you know, personnels can turn it around for today, I did a tax loss harvesting and I'm selfishly hoping that, you know, that insurance news, you know, comes after my 30 day period when I can revisit the stock.

Obviously for patient's sake, I do hope it comes as soon as possible. Um, but you know, I, I'm wondering, you know, these stocks, these other stocks, Tempest and Natera have been weak. And yeah, there's been, they cited basically that pharma tests and maybe it has to do around these tariffs.

I, I listened to the call, but it kind of felt like vague commentary there and they just said, you know, volumes were down. So I'd imagine they're probably not the only ones feeling that. And we'll see in the next couple days what these other guys report.

But, uh, I also dumped my tempest position for now just 'cause I got, you know, hit on this PSNL and you know, I'm gonna wait and see what happens. But anyways, uh, let's just, I wanted to get those comments in, uh, just for what happened in this week. We can call it there.

Uh, yeah, it was a great show. Guys. Anything else, Shay Tune in next week for episode three of season two.

Can we just end the earning season now? Like, this is just crazy, man. I know.

I, I want, I want sleep. I know I want sleep. Let's take A break.

It's like this morning I was like, oh, great, I, I don't really have that much reporting tomorrow. And then I saw Oscar's reporting. I'm like, oh man, here we go.

There's a lot more Woke up. I woke up Tuesday morning thinking it was the weekend. Oh, what a nightmare.

That's, That's, that's how exhausted Monday was. It's me eight, Dude. Really good.

Four Names today, man. Four names. I feel like, you know, when you're in college and you had like three finals on the same day, that's exactly what I felt like.

Yeah. Oh my goodness. Don't remind me of those days.

And then you had to, you had to study for one of them. You had to put all the time into one of them. You couldn't do all of them.

Anyway, guys, was a great episode. Cap up in episode two of season two. Uh, definitely, uh, keep an eye out for episode three.

Coming out next week, we have way more earnings coming out. We haven't even really crossed into the bulk of software earning season yet. That is usually a follow up of the, uh, major me mega cap earning season and maybe mid cap earning season, but it's not even close to being over.

And also the fifth Super Bowl of the year, or in this case, this would be the third Super Bowl of the year. Nvidia earnings is coming later this month, so that's probably gonna be something of a discussion episode five. We'll see.

But, uh, I once again guys, thank you for watching. Thank you to future equities for sponsoring this video. We're hosting this, producing this video as well.

Appreciate you guys. See you guys in the next one. Take care.

Hey everyone. Welcome back here to Techstrong tv. You know, my, my next guest reminded me, I think the last time him and I were face to face.

Might have been at a enterprise, a DevOps Enterprise Summit does, I believe it was. Yeah. Which That's gotta be maybe before COVID Pete, I think it was.

I think maybe 2019 was the last time you and I Talked right before, So, Yeah. Yeah. Hey, it's great to, I don't, you know, I'm sorry it's been this long and it's great to have you on here on Text Trunk tv.

I hope all's well with you. How are you? I'm doing great, and I'm really happy to be here.

You know, my co-founder, Robert Reeves, used to, um, do a lot of the forward facing stuff. They're getting me into more of it now. Yeah.

So we, we will probably be talking more as time marches on Good Abs. I hope so. I hope so.

I was, as we were talking about off camera, I just saw Robert. I didn't see him. I did one of these, a Zoom call with him, which I guess I saw him.

I just didn't see him in person, as they used to say. Sounds outdated, like recording stuff on film. Anyway, um, Pete, I, you, you mentioned you're a co-founder of Liquibase.

Give people a little bit of your journey and, you know, and the, and we'll segue from that in a liquid basis story. Sure thing. Yeah.

So I was born and raised in Austin, Texas. I've been a part of the tech scene here since the late nineties. Um, right around 2000, right, right before Y 2K.

Uh, and over the last 25 years, or 26 years, I guess at this point, I've worked at several startups that got acquired by larger companies. There was a logistic startup that got acquired by Neopost in Europe. There was a security startup that got acquired by Symantec.

Um, and then prior to starting Liquibase, there was a, uh, a another DevOps company really, um, as DevOps was kind of taking shape and, and becoming a movement called Furnace that managed J two Double E cer configuration. That company got acquired by BMC, and it was really our work with BMC that led to sort of the seed of starting a company that does what Liquid Base does now. So we were, man, we were helping people manage their JTW server infrastructure a lot more easily with automation, you know, reducing all of the manual errors, allowing it to be repeatable and automated.

All the great things that, that you really want from sort of a point solution and DevOps. Um, and as we were doing this, we heard from some of the largest brands in the world, Hey, this is great. This is gonna help us.

This is how it's gonna affect our bottom line. Do you guys have anything planned for the database? And it wasn't just one or two people.

We heard this again and again and again. And as we kinda started probing into those conversations, what we did, what we discovered was while all of these things were being automated around the database, the database itself wasn't being automated. So you had tools like Jenkins, uh, starting that ci cd journey for a lot of companies.

All of the automation around testing, building, assessing, security, producing an artifact, all of that was being written for the application code, no problem. But people weren't doing the same thing for the database. There were no solutions for the database.

It was still a largely manual process. And, you know, not only was it risky because it was a manual process, but it was slowing down the investment. Everybody was making an application code because you can't really ship an application without chipping the database changes with it.

So, um, we looked out in the world of open source, uh, as part of our due diligence and our initial market validation for starting this new venture. Um, and we discovered Liquid Base, the open source project that next year will turn 20. So it's been around since 2006.

Uh, and, and really it exists to help developers organize, repeatably, execute, and automate database migrations in conjunction with application code migrations. And that's what we live to do more and more as time has marched on, it's less about actually automating those things, but providing the, uh, guardrails around that automation. So making sure that you're, you're maintaining security standards, making sure that you're maintaining audit and observability standards, making sure that the changes that you're introducing to the system are of high quality prior to being merged into the code base.

So a a lot of work on the developer front, um, for the developers who are now managing these database changes to really provide the guardrails and the assistance that they need to do this well and make sure that they aren't introducing, you know, performance or security issues, uh, into their database. Um, and that's what we've been doing for the last almost 14 years. The company will turn 14 next year, next to April.

Um, so yeah, I mean, that, that's, that's really why we exist. We want to help you manage your database changes with the same level of visibility, security, safety, and, and, um, efficiency as every other, uh, component of the application stack. Absolutely.

com in, uh, 20 13 20 14, we first published early 2014. Um, of course, liquid Base was known as data at the time, right? Yeah.

Liquid based was the open source project. But really, you guys kind of pioneered DevOps for databases, because you're right. Databases from we're kind of the redheaded stepchild of, of the whole IT stack, right?

You know, the, and the, I don't know why. Well, I do know why. 'cause I've seen it over the last three or four years where data has become pri primacy again, right?

That data is, for a little while there, early two thousands all the way through, let's say 2020, we seemed to have laws. It felt like the Bill Clinton campaign. Remember, it was keeping, it's about the economy stupid.

We remembered that sometime around 2020. It was about the data stupid, it's about the data. We could put all of this effort into our application and our application processes and and development.

But really it was about the data and, and the databases, of course, the container for our datas. And, and so there's been a refocus on database and the data mm-hmm. The primacy of data.

Yeah. So, and liquid base is wr rid ridden that kind of wave now, right? Or rode that wave.

Yes. I guess is the right word. Yeah.

Yeah, absolutely. So, so that's, that's kind of what, what, you know, over the course of the company we've noticed as well. And I think that 2020 date is really kind of an important one because I think that's when we really started to see the initial, um, sort of get a hazy picture of what AI was gonna do for us in the future.

And obviously, you can't really train these LLMs, you can't have, um, robust, um, you know, robust LLMs that don't hallucinate as much without high quality data and understanding the lineage of that data. It's only As good as your data. Yeah, yeah, Absolutely.

So, I mean, I think what we saw is, initially people were trying to fix a functional problem, but now they're trying to fix a strategic problem. They're trying to stay competitive in a world where AI features and AI tools are launch popping up like crazy. Um, and, and, you know, there's a real sort of, I think there's, there's a pretty, um, uh, uh, sharp increase in the cadence with which people feel like they have to move in order to stay competitive.

So, um, you know what, we found our customers that, you know, have been customers for several years and kind of have the operation of managing databases on rails in their automation already. It was really a lot easier for them to start, uh, you know, launching these AI initiatives because their data was already, you know, there was a high level integrity because it was being consistently managed across the pipeline. Um, and, and those were the things that kind of are, are driving our customers now.

It's not so much about the initial promise of DevOps, but it's about supporting these, these AI initiatives that are really just kind of have spiked the demand for high quality, um, understandable data. And, and that's, those are kind of the problems that we're solving with our customers now. Love it.

I love it. com? com is the best place mm-hmm.

To jump in and, and start learning about Liquid Base, familiarize your yourself with Liquid Base or see what's new in Liquid Base, if it's been a while since you've checked it out. Um, we've got, from there you can launch to, you know, our GitHub repo. We can get to our documentation.

We have a whole site on contribution and, and things like that. We have forums where, uh, that are pretty active, where our users go to discuss usage of liquid base and troubleshoot issues. com is your gateway to all of those things.

So I, I would definitely start there if you wanna learn more about Liquid Base. Excellent. All right, Pete, I'm gonna shift gears a little along here.

I wanted to talk about, well, it's actually what Liquid base calls the database delivery gap, fixing the last mile of DevOps. And a lot of this comes out of a, uh, uh, report insight report. You guys did the 2025 state of database DevOps.

Um, actually before we even start discussing, for anybody who wants the report, they can get it, I'm assuming from the liquid based site. Do you have like a friendly URL to give them or just go to the front page? I, it, it's on our blog.

So I go to liquid based slash blog and just look for state of database DevOps. com/b/state of database DevOps report dash as dash of dash one h dash 2025. But it's probably easier just to go to the site out in The blog and look up state of report.

Exactly. All right. Um, spoken like a real database person there feet.

Um, so talk to us, what, what's in the report here about the database delivery gap? Yeah, so really we, we kind of touched on it earlier. You know, the more things change, the more things stay the same.

There is, uh, because there are, you know, factors like AI and really just, you know, some, some of the, uh, the later companies in, in the process of adopting and, and building out their DevOps practices. Um, you know, it's really driving this focus on the database because there are a lot of, of, of proven tools and patterns and things like that for everything. But the database changes.

But kind of getting the database changes to move and lockstep with those database changes into benefit from the same automation is really what the biggest gap is for a lot of our customers. We'll see folks who have homegrown solutions, or they're using an ORMA framework of some kind, but frameworks don't really provide all of the, um, necessary governance and observability benefits that you, um, that you really need in this day and age of ensuring that your data's in, in high of high integrity and that you're satisfying audit and compliance, um, needs. But it's, it's really what we're seeing is that, um, you know, not only is there a gap in process, there's also, uh, kind of a gap in, in understanding and training.

So, um, yeah, just to kind of cover some of the highlights of that report, obviously you were talking about this a second ago. AI and ML are really driving sort of, you know, an explosive, um, need for data, a demand for data that unlike we've ever seen before. We need more of it.

We need it faster. We need it, uh, to be of high integrity, and we need to, um, make sure that we are managing the process of moving it from our transactional databases to our data lakes or our S3 buckets. Wherever, wherever we're analyzing our data, we need to make sure that process is on rails.

So AI is really kind of driving a lot of that adoption. And this was, this report came out of a survey. We surveyed, you know, a lot of people, customers and, um, OSS users alike.

Uh, and, and in that we found that 78% of those respondents said the biggest challenge they're facing right now is rising to the challenge presented by AI and ml. So, um, but what we also noticed is that the teams who had been doing DevOps for a while and kind of fell into our mature cohort, um, it was really just kind of business as usual. They already had their databases on the rails.

They had all of the audit and lineage information they needed. They were sure that their processes were fluid, they were guard rails to keep them from breaking. So really it's just like, okay, we've got more data that we need to ingest and move from A to B once they have database DevOps sort of, you know, baked into their current processes.

So that's a lot of what we're dealing with right now is com. Those, those companies in the 78% that are trying to get there more quickly, they're having a real tough time with that. But with this new proliferation of, of tools and, and, you know, processes and different data platforms and different patterns for managing that data, um, you know, we also found that a lot of the practitioners feel like they, they're not really getting the training.

They, they need. They're being given the tools they need. They're being given sort of the initiatives and the, the, the milestones they need to hit.

But what they're not getting is, is training or, or advice or, or really just kind of that first step in DevOps, which is, you know, let's understand our entire process end to end so we can all bring our different perspectives together to build the best possible process in terms of speed, safety, and reliability. So, um, yeah, I mean, it it, it's really just a lot about, you know, we're seeing a lot of new tools, we're seeing a lot of new patterns, we're seeing a lot of great innovation in the space. But I think that, you know, what we're hearing from, um, survey respondents, what we cover in the report, a lot of people feel like they aren't getting sort of, we aren't investing enough in the human aspect of that, the humans that use the tools.

So I thought that was really interesting as well. And that's something that I've been talking to people about, um, you know, pretty consistently whenever I get the chance to speak to them, for sure. I love it.

Pete. Sounds like a great, a great report. We said where, where it's, uh, available.

What else going on at Liquid Base? You guys gonna be at any shows anytime soon? Or maybe CubeCon, I mean, or something like that?

Yeah, I think, I think we'll have Boost on the ground at, at CubeCon. We're looking at, uh, reinvent also, and we're, we're also mm-hmm. Um, I moved into more of a developer relations role this year.

We're focusing on local events as well, so various, Well, Austin's a hot area PCDs or things like that. So yeah, I mean, I think, we'll, we'll be around, we'll be pretty visible, um, at, at the major meetups. And then we're doing some regional stuff too.

So I would love the opportunity to talk to everybody, anybody that wants to talk databases or DevOps or really anything. Um, and I'm pretty accessible. com.

I make it easy. I love it. Steve, it's been a while.

Don't, let's not wait five, six years for you to come back on Text Trunk tv. Okay. Agreed.

Agreed. Yeah, it was great. Great talking to you again.

I look forward to the next one. Alrighty. Pete Pickerel, co-founder Liquibase here on Tech Hor tv.

We're gonna take a break. We'll be right back. Hello and welcome to the latest edition of the Techstrong AI Leadership Inside Series.

Today we're with Dan Fernandez, who's vice president of Product for developer Services at Salesforce. And we're talking about, well, the rise of vibe coding and model context protocols and all kinds of fun stuff. Dan, welcome to show.

Yeah, welcome. Thanks again. Happy to be here, Mike.

Yeah. For the uninitiated, what exactly is vibe coding? 'cause some folks would just say it's sun of no code slash low code with a little AI thrown in, but maybe there's more to it than that.

Yeah, you know, it's kind of funny. It's, uh, depending on where you stand, you might see it as almost a derogatory term. Like, oh, I'm just, I'm just describing this stuff.

And there, there's sort of no, no care or thought within here, but I think it really is sort of thinking about, uh, the requirements of what you wanna build and making that the focus and letting AI really do a lot of the rest. And, uh, one of the key areas is obviously making it an iterative process. How do we sort of understand the requirements, build a plan, and then iterate on those requirements to understand?

And it really is faster time to value. That's the number one thing we're really trying to drive. But that, the, the downside, well, it's insecure.

Like we have, there's a number of stories where we see, uh, folks releasing things and things like personal data driver's license. You don't necessarily wanna have somebody vibe code your, uh, medical history, right? So when it comes to enterprise, uh, vibe coding, uh, you really want to think about security and within the context and how do you apply sort of, uh, uh, policies and governance, uh, so that you're building safe, secure applications.

And that from an enterprise perspective, you can delineate, uh, that process. And that I think is really sort of the, the difference between vibe coding, and I think the, the latest area that's really taken off it's model context protocol, which is the, uh, you know, it's sort of thought of as the, the USB for being able to plug in different services. But what it's doing is, hey, if AI or, you know, an LLM let's just pick on Claude can, uh, generate, uh, application artifacts, instead of just doing application artifacts, it can now take actions on your behalf.

So call that API call that website, do that specific action, run the query, visualize the results, and that's something that has really sort of taken it to the next level. Hmm. And vibe coding, is this something that I as a, say, citizen developer or doing?

Or is it basically just something I'm instructing a bunch of AI agents to go do on my behalf? Yeah, I think that's one of the, the real areas is who is this for? Uh, and there's sort of two answers, which is, hey, how do we democratize development so that a bigger and broader audience of folks can participate in that?

So that's, uh, maybe it is more product managers, UX designers, analysts a business, uh, people that own a specific, uh, business could literally describe their application and generate that, that application. So what are, like, a classic example might be what you'd almost call personal apps, which is meaning I'm doing it to solve my problem, to answer some question, which is, I'm not planning on intending to share this publicly, but I gotta write some complicated data query and I can copy paste things into Excel and, and spend a bunch of time looking at vlookup. Or I can ask an LM to basically build that visualization for me, or look at this CSV and tell me what's interesting.

That's a great use of data. Uh, and things like Tableau allow you to actually visualize that, show me that in a bar chart, show me trends over time. So there's great capabilities on almost the personal apps.

And then prototyping is sort of another killer app as well, which is, Hey, I don't really know what we want yet. We wanna just meet with customers and let them see the application, get feedback. And so instead of, you know, getting in the backlog of it, we can get feedback, get a better understanding of the requirements, get a better understanding of the user experience.

Um, as just kinda one example that we've been working on is what we call Figma to LWC, which is, hey, our UX designers, uh, product manager and UX designer are working in tandem. Product manager has some ideas. They, they don't necessarily have to wait for engineering to build that prototype.

They build the prototype, it's a working prototype, it's not gonna be the finished product, but allows you to get feedback and the designer's able to work from where they want to within Figma. And then you take that Figma design and literally build the application based on top of that. So it's sort of that great, uh, uh, how do we basically increase the feedback loop for building a better product?

So we didn't spend three months building the wrong thing. So will this reduce the tension that often exists between so-called citizen developers and the professional developers out there who frequently find themselves trying to iterate some sort of application in a low code environment only to have to go back in and kinda rewrite it in at a, at a lower level language so that it can scale and run and qualify to run in the enterprise, but it's the dynamic between those teams gonna change. Yeah, I, uh, I think so as well.

Uh, just to sort of take a step back, when you think about, uh, enterprise ai, it's how do we think about having a great application lifecycle management, right? So like, hey, we're building these applications. It's not like, um, that citizen developer can just take it and go with it.

How do we define, uh, what that governance is going to look like for those applications, which is, hey, this is like the bar. And we, uh, for example, recently launched a product in DevOps center testing that allows you to define quality gates. And, uh, the example I kinda use is federal, state, and local.

Meaning like an enterprise may wanna set all applications, must follow these specific rules, and maybe there are security rules. We're gonna have code analysis tools. We, uh, have a code analyzer tool that basically will go through and find, you know, things like security issues, best practices, scale issues, uh, uh, SoCal injection, things like that, that you wanna make sure that you're protecting yourself from.

So everybody must follow those sort of federal rules, if you will. The state rules might be, uh, um, uh, within an enterprise, HR or finance wants to set different rules for what their applications are. Maybe it's auditing and compliance rules that they must follow so they can set the governance.

And that includes everything from, uh, data governance, uh, role-based access control to even AI rules, which is we've built a bunch of, uh, ip, we don't want our citizen developers to basically go and reinvent that. We want you to reuse that ip, which is, we have a set of services, we have a tax calculator, don't rebuild the tax calculator. We have a formula for, uh, finding out discounts, and this is how we set discounts, reuse the existing IP we have.

And maybe that's built by proco developers. Uh, and so the, the barrier to entry becomes that much easier for reuse across your enterprise. That's one of the challenges with the public vibe coding tools.

They have no context of your enterprise. They don't know that you already built a ca tax calculator, shipping calculator, or you have these backend services, or you already have a representation of a customer, a formula. Uh, that's one of the key differentiators with both, uh, uh, enterprise context and MCP that allows you to reuse those services and get as much of your existing IP that you already have.

So really, the, the citizen developer becomes more like, I am creating glue. I'm making sure that it passes, uh, all, uh, the requirements for enterprise, and then I'm going through an application lifecycle management to make sure that any updates to that application are following what constraints we have for development testing. And again, so that we have, uh, high quality and, and, uh, observability for when the application goes out as well.

Mm-hmm. You mentioned MCP earlier, it seems to me that that's a big part of the glue we're talking about that makes the data accessible to the vibe coding tools, but, um, that can be both a good and a bad thing. So do we need to apply some sort of governance around how all this stuff gets used?

Yeah, that's exactly right. So, uh, for example, we launched the Salesforce DX MCP, uh, it's available, you know, it's getting like 10,000, uh, downloads a week on, on NPM. And it allows you to basically do natural language questions for anything.

How many administrators does this specific org have? Show me this particular data in this particular way, all using natural language, it converts it into a query and, uh, and obviously you can build applications with it as well. So, uh, incredibly powerful, but it does all based on the foundation of, uh, the role-based access control that we have within our tools, which is setting things like permission sets.

So it's not like you can just access data, it's using the context of the currently logged in user using, you know, not to get too technical, but the, the JWT, the jaw token. So, uh, Dan can access certain amount of data, but, uh, I can't access all, like, say, HR records or, or healthcare records, right? It's using the context that I've specifically been given in, uh, uh, so we're not having that issue of, of those security issues.

And that's all because we're building on that deeply unified platform that Salesforce has, right? So it makes it that much easier. You don't have to worry about the scale or the uptime and everything else.

Worry about your business logic will actually, uh, help you do tools to BI plan build and make it have quality by default with our code analysis tools and our testing tools. I think the other great one, uh, that we have is, um, a tool called Data Masking that's available within our sandboxes. And you can think of this as, Hey, how do I test this MCP?

How do I make sure that things are safe? Uh, you run your application tests in a pre-production environment. Uh, we have one called the Full Sandbox, which is all of your data, except it's been anonymized with data mask, right?

So it's not actually my phone number, it's not my personal identifiable information, but I can then run and make sure the application works when, not just when I have a hundred rows, but when I have, you know, a billion and really thinking about the end-to-end application lifecycle management, uh, tool for that. So, uh, that makes you basically be able to have trust that you can verify, uh, the MCP tools aren't doing things that they shouldn't be able to do. Mm-hmm.

Um, to your point about the types of applications that we're gonna build, uh, you hear people say, we're gonna build more apps in the next few years than we built in the last decade, but a lot of those apps are gonna be, as you described it, personal apps, right? So Correct. How do I kind of distinguish between what should remain personal?

'cause a lot of the end users, they say, oh, I have this cool tool, and they start passing it around, and the next thing you know, it's an enterprise app, Uh, that is exactly it. And it sorta is, uh, uh, I don't know, uh, it, to use the, the scientific way to think about it, but it's almost like evolution, which is, hey, if that app really becomes and solves a problem, this salesperson did this one thing and it visualizes, uh, their top prospects and what are the great things they have? It uses transcription services, audio transcription for meetings and, and automates a bunch of the work they have, they show it to one other person and it becomes sort of viral.

The, if you started with that enterprise policy and governance, you can assure you like, Hey, whatever that application built, we can make sure that it was always, uh, uh, building on that foundation of security, right? That the sort of federal rules, if you will, if you just say like, Hey, everybody can do everything else, then that's the problem where it comes in and Wow, this just became, now I have to spend a bunch of time making sure that it does follow the, the minimum requirements and everything else. So it really is a good idea for us to think about that policy and governance upfront and what are, uh, what are the, the things in the dials that we will and won't allow, uh, uh, our citizen developers to do.

And, um, that is one of the key areas that we're seeing. It's when people don't have that policy or don't start thinking that forward looking is the application and oh geez, we're gonna have to re-platform this. This is using different data.

It's actually, you know, uh, doesn't have SOC two compliance, you know, you should have used our official data source and all that fun stuff. That's really where we, we get into the problem. So if you start thinking about that governance and policy upfront, uh, it makes things that much easier.

Mm-hmm. Most of the professional developer organizations I know have a significant amount of application backlog work that they're supposed to do. That's exactly.

And then look through that and maybe kick some of those things back to the citizen developers and say, you know, go use Vibe coding to build this thing, because it doesn't really rise to the level of effort required for professional developers. Yeah, that's exactly right. I mean, think of many departmental apps, like, I just need this one thing and it's always gonna be the priority 15 for the organization.

So how do we enable folks? It's like great marketing person. You basically wanna have some way to collaborate and just share, uh, assets or artifacts, logos across different marketing teams inside an or a large organization.

You can absolutely, uh, uh, build your application and empower those folks set up sort of that fundamental, uh, policy and governance and set up their dev environments and test environments so you can make sure before anything gets to production, Hey, you have passed. You must be this tall, if you will. These are the quality gates we're making sure that you run with our automated code analysis tools and, and things like that.

And that really is the key, but it is a great chance for us to really lessen the load and also to have that conversation, what really is important for it to only do, or that we can empower and democratize, uh, the right, uh, folks across our organization. And that will, uh, uh, lead to like good conversation. I'm like, actually, no, this is critical.

We definitely want it to do it. Or Hey, uh, let's empower that marketer to do it. And there's certain parts of the application that we wanna make sure anything within Data Act, we're gonna build an API, we'll expose that as MCP and, uh, you know, it enables the B2B website and all that other fun stuff.

So really thinking about what are the intersections of where we want those things to do, sort of the, the plumbing and piping that we want, uh, uh, different folks to be able to do and setting those across it in different departments. So is there a smart way to introduce this, or do I just kind of give everybody the tools and we all experiment on each other until we get to something that works? Yeah, I, it's a great question, which is like, Hey, how do you do it?

And this is always different with different, uh, uh, organizations in terms of like, Hey, how do we set, how do we start thinking, uh, about this as just kinda one example, A number of folks just start with the pilot. Hey, we have one team. Let's just take one of those applications off the IT backlog.

Let's see what would happen, and let's figure out what those guardrails are. It turns out our IT team already has a bunch of unit tests and they can already use our automated code analysis tools, no additional costs, and we already have pre-production environments set up for them. Great, let's go do that.

Focus on the app and see what we learned, what worked, what didn't work, uh, where do we sort of separate, uh, the boundaries between, uh, uh, the citizen developer and the PRO code, uh, IT organization and think about what, uh, that is. And that's probably the best place to start. And it's like, okay, based on that, here's how we think we could roll this out within an an enterprise.

We wanna get, you know, uh, a hundred teams inside, uh, we're gonna do, you know, recorded training and really think about putting AI in the hands of folks. And it really is like the, uh, it's almost like fight Club. Like, Hey, we're gonna show up for a meeting and we're, you are gonna build your first app, and here's what you're gonna do.

Because, you know, our, our default project is set up with that enterprise governance. You can have this sort of confidence within there, but describe what you wanna build for your first app, get on guardrails and really set up a way for them to feel empowered, but also, uh, uh, that you have the, the knowledge and secure, uh, of the safety and security, uh, boundaries that you've set up within your system. Hmm.

When we put all that together too, at some point, all these apps are gonna get interconnected in some ways, and we see people talking about, um, you know, things like the agent to agent protocol, but do we need more infrastructure beyond the MCP to kind of make this whole vision turn to reality? Um, it's a, it's certainly a good question. Like, uh, we have some examples that are incredibly powerful today, right?

Like the 1-800-ACCOUNTANT, uh, which is a service. So you're an SMB, you don't really wanna be an accounting expert, right? You're focused on, you know, your floral shop.

The floral shop goes from one shop to 15 shops. You don't wanna do the accounting of it. And so, uh, they, they were able to set up, uh, 70% of their inquiries were actually solved by, uh, an agent that they set up and literally pointed it to their existing knowledge base, right?

So fast turnaround, uh, and is able to take actions based on their behalf, answer basic tax questions or even simple, simple things like, Hey, when do I get, uh, password or resets? What's my return status? And so on and so forth.

What's sort of like the next level of that is really interesting as well, which is, hey, we are starting to see it's not just sort of knowledge base agent, but actually action agents. And maybe it's more like, Hey, let me walk you through and really think about sort of that personalized experience. And maybe that agent is talking to something else, which is, Hey, we are connected up into a bank, US Bank, Wells Fargo, and it's actually doing some stuff based on your data.

I may that you were able to interconnect with me, I'm able to suggest better options for you. Hey, you bought this thing. If this was a qualifying purchase, you could actually take a tax write off on it, for example.

That's an example of like, Hey, how could we actually get better knowledge, uh, within MCP? The other area is even MCP is evolving to support things like elicitation, which is a fancy way for structured input, which is like, Hey, anytime I'm gonna ask you, uh, for certain things, tax codes, I need to understand like where your city, state, and your county to be able to understand that developers are now able to build those elicitation to always make sure to have those. And I think you're gonna see that that's for input.

You're gonna see that for output as well, which is, Hey, do you want me to visualize this data so you can see your tax returns the last three years as well? And then it gets a little wild when you start talking about those agents to agents. So maybe, uh, uh, you start thinking about, Hey, I'm going to have a set of folks, this is my, uh, uh, accounting agent is talking to my inventory agent, which is talking to another agent within there, my storefront agents, and they're all collaborating and actually suggesting path forward.

I think the key though is there's always humans in the loop. Uh, like how do we make sure that we're thinking about, uh, it's not just a agents kind of going off, it's really what is the problem that we're trying to have them solve? And being able to have that federated control of information as well.

That'll be another area where it's like, Hey, I don't want my agent to leak information to another agent. I wanna explicitly set, uh, uh, the controls for how I'm exchanging information across those agents. All right, folks, you heard it here.

Usually the best way we learn is we play. And once we start playing around with stuff, then suddenly vibe coding doesn't seem so weird and different, it's just the way we're gonna work. Hey Dan, thanks for being on the show.

Yeah, this is great. Thanks again. All right.

And thank you all for watching the latest episode of the Techstrong AI Leadership series. You can find this episode and others on our website. We invite you to check those all out.

Until then, we'll see you next. Hi everyone, thanks for joining us again in this breakout session of the Red Hat Cloud Fridays with AWS event where you get real talk about real solutions. This session is going to focus on Red Hat Enterprise Linux, and our title tells you that we are looking at how to go beyond standard and tell, unlock the real value of RL to get value on AWS.

I'm Simon Briggs, I'm the ecosystem sa focused on AWS across Europe, mainly Eastern Africa at Red Hat, and I brought a special guest with me today who will introduce him in a second, but it'll become plain that he's a real expert on the power of RL and will present the majority of this session. So if you'd like to introduce yourself. Yeah, good, Good day everyone.

My name is John Van Breen. I'm a, um, specialist solution architect looking after our Relapse Enterprise Linux platform. And thank you for having me on this panel today.

It's good to have me on board. Let's dig into what we're gonna talk about in this session. So we're gonna talk about why AWS cloud is important, um, as a channel for customers to be able to get to Red Hat, um, software.

Um, then we're gonna talk about the different ways that you could buy Red Hat Enterprise and I send some of the other software. I'll just, um, mention it in passing and then we'll get to the meat of this presentation where, uh, Yuan, who I can never pronounce his name properly, so I do apologize, uh, will go through the power of Red Hat Linux. Now I have, um, mentioned that q and a is encouraged and you can use the chat section to enter questions throughout our session, but we are actually going to answer any questions in our conclusion session after this one, which will run as part of the, um, cloud Friday event.

So please ask away. We encourage you. So why is AWS really important?

Well, the fact is Red Hat understand that our customers love to work with AWS and of course, when our customers need to do things, red Hat listens and this has been backed up with research. So, um, last year, 2024 Red Hat, uh, worked with Canales, another name I can hardly pronounce. Um, and producer report when we were analyzing how important ISV or third party products are, um, when being purchased via the hyperscale clouds.

And this is an absolutely massive, uh, market that has essentially grown up in only 10 years since AWS brought out the first genuine, um, hyperscale marketplace. Um, it's looking at in 2023, it was around $16 billion. The forecast for this year, extrapolated from that research goes through to $45 billion.

And as you can see, the expectation is that we'll only keep growing. So in tha 2028, we are moving towards $85 billion. And in my personal experience, I joined Red Hat to work on, um, hyperscale cloud partnerships because I know that any of the analysts expectations are conservative when you look at the explosion of interest within the industry about utilizing the services the hyperscalers can provide.

Importantly on the right hand side, there was a couple of findings in this report that I wanted to bring to your attention. First of all, um, many organizations are looking to use a hyperscale marketplaces to involve third parties trusted advisors effectively in that procurement process for software. And that actually is an interesting capability that we've now got available if you are looking at buying r through one of your partner organizations.

And also, sometimes we found that organizations didn't get the best commercial, um, arrangements wrapped around their investment on third party software coming through this route. Obviously this is a survey from a year ago and organizations are maturing all the time, so things will have improved, but back then we saw there was lots of dollars being, um, left behind by customers that they could have accessed through buying programs and funding. And certainly some of the capabilities we have in offering Rail to you will help you access better buying conditions.

So saying that, how do customers buy rail on AWS? Well, there's multiple options for you, but at the core of all the options is Red Hat's understanding that when customers are wanting to buy through AWS they're looking for a cloud experience that first of all gives them flexibility. Our customers constantly tell us that they want to be able to buy efficiently where they need when they need it, but at the same time, there needs to be quality and the stability of the service that they're purchasing, the capability that they're purchasing needs to be solid to be able to drive the critical business applications they deliver on them.

And also they need to be confident that there's a secure wrap around that technology. Of course, security has always been pri primary factor in customers understanding of how they should procure software for their businesses, even when they've been working solely on, um, on-premises delivered platforms. But in the modern world, extending into cloud services as well as including those, um, on-premise platforms, um, means security is even more critical to organizations.

And certainly we will touch on that in our discussion about how powerful Red Hat Enterprise inex is. And then also touching on that affordability piece where we mention that in our analysis some organizations are not getting the best buying conditions, we understand that we need to make that available to customers. And because of that, we have several options available to you as an organization looking to buy Red Hat Enterprise Linux today.

So there's four different motions that you can, um, use to be able to procure the software. And let me, um, describe them in this way. The two outer options here, the two pillars on the far left and the far right, um, are, um, routes to procure software that have been available for many years on AWS cloud.

Um, the one on the left has always been available, so customers have always been able to buy Red Hat Enterprise in it for all the powerful reasons that we will discuss later in this session from Red Hat and our partners. Um, and then take it themselves and utilize the subscription they've purchased on AWS cloud. The right hand column, the console column describes another buying motion that came about a few years after.

Um, EC2 became a standard service from AWS and that was where Red Hat and AWS worked together to turn AWS into a Red Hat certified cloud service provider. And in that case, AWS sells Red Hat Enterprise Linux baked on top of EC2 images to customers on our behalf as a cloud service provider. In that scenario, the customer is able to procure in a PayGo fashion from AWS.

But a couple of things, um, are, um, important to understand when customers buy that route, buy from that route. First of all, AWS has a fixed price in that model. We work with our cloud service providers and, um, those service providers provide a royalty back to Red Hat in the background for the software that's used that is ours.

Because of that, our pricing is quite fixed and there's no scope to be able to buy, provide any special pricing to our customers. Um, although importantly, when customers do buy PayGo via the console, they are buying through a cloud, um, motion. So therefore it is via AWS as a partnership and it's recognized to spend against any spend agreement the customer has, whether it's an EDP or a private purchasing agreement with AWS.

The spend on the Red Hat software on top of the EC2 is recognized. As I say, it's a cloud-based transaction, but you don't see that this is, um, channel friendly. So where we were discussing, um, earlier our research about the use of, um, independent software vendors services on AWS, many organizations were using partner organizations to, um, e exploit the value of those services.

Um, in the case of the console procurement route route via AWS, the um, channel has no place and that's why the two offerings in the middle have been created and recently announced. So Red Hat now has the ability to sell Red Hat Enterprise Linux via the AWS marketplace. Importantly, you see at the bottom that, um, that opens up a channel friendly route for our customers.

Um, the, uh, business value adding partners that they work with can, um, help in that procurement decision for the customer. There is a place for them in that, um, procurement route whilst it is still a cloud-based transaction. And the spend is recognized in a, um, against a committed spend with AWS as well.

There's two pillars 'cause there's two forms of consumption from a customer point of view. They can, um, either, um, subscribe to a committed term with Red Hat via, um, a marketplace offer, or they can commit to a wider consumption deal with Red Hat that covers multiple Red Hat software products. If you are interested in the consumption model, please speak to your Red Hat representatives or ourselves later on in the Cloud Friday event, and we'll be able to explain more detail about what, uh, a consumption agreement with Red Hat will look like.

But rest assured there are different options for organizations importantly with the marketplace offers that are available because Red Hat is the organization selling to you, you are actually buying via Red Hat, but procuring through the AWS marketplace, we can offer through private offers, um, extended pricing, um, agreements, which would recognize longer term commitments to the software packages and also the amount of commitment you spend with Red Hat, um, in entirety rather than the PayGo model I mentioned earlier. So what actually are we talking about when we are talking about those different options? The option when I talked about console is represented in this AWS slide.

So you can see there that many of the Red Hat Enterprise Linux server derivatives are available to today, screw that route. But when buying through that route, you are buying directly with AWS. We have the ability to, uh, deliver SAP with ha.

Uh, we have the ability to just deliver Linux server with ha you can deliver, um, across many of the instance types that A-W-A-W-S support, be it graviton instances, uh, outpost instances, arm instances. And we do have special builds for application types such as SAP or Microsoft sql. But, um, on the right hand side, we've got a more extended list of the software you can procure on the AWS marketplace.

You can still buy the types of software I've discussed as being available on the console. It's just different buying motion. So you can still buy Red Hat Enterprise Linux in the forms we've discussed for console, but importantly there's one extra, um, uh, uh, type of rail offering in that marketplace list.

The Red Hat Linux, including third party Linux migration and extended lifecycle support. That's a special kind of listing that Red Hat has made available via AWS. You can buy it direct from Red Hat yourself or you can buy it via the AWS marketplace.

But that, um, subscription allows organizations to take any derivative, um, operating system that they have today that maybe is out at support such as CentOS and migrated across to be a fully fledged Red Hat Enterprise Zenex instance to get full support and migration supported from that subscription. Um, I'll also highlight that recently we're very proud we released the Red Hat Enterprise UX AI listing on the AWS marketplace. That's a Red Hat enterprise ux, um, instance, which is specifically curated to give value for organizations who are looking to build trial and deploy their initial AI workloads.

So if you are looking to build agents or bots, which will add benefit and features to your organization, but are not yet at the point of wanting to deliver highly sophisticated collaborative ai, um, infrastructure across a much wider platform, then Red Hat Enterprise Inex Server is a fantastic solution to help you access those capabilities. It gives you access to GPUs and also a curated version of the Instructor Lab's open source project, which allows you as an organization to augment your large language model with your specific, uh, intellectual property and data sources to be able to deliver a particularly customized AI solution to your organization. So that's a quick high level description of what Red Hat Enterprise inex services are available today through AWS.

And at this point we'll change gears and allow our specialists to talk to you about the power of Red Hat Linux and how it works on AWS as a cloud. Thank you, Simon. So you've explained, you know, the options you have in terms of getting greater Enterprise Linux or Rails is more commonly known in into AWS, but Red Hat Enterprise Linux is probably the most well-known Linux distribution out there, and we have thousands of customers that's using it within the cloud environment, but also on-prem.

And one of the key things that we've trying to find is that very few customers only run in an environment such as AWS typically have, um, that on premise, they may have it in some of the other cloud environments, they might even have it in edge devices. So Simon, one of the key things that we're looking at is that customers aren't just running the environments singly on a AWS environment. I have very few customers that's sort of in that scenario.

Most of them have been migrating from the on-premise environment into cloud environments. They may have some edge devices and so forth. So this is becoming a very complex environment for them to be able to manage and look after and how do they get that flexibility back into it to be able to see what is going on, to be able to quickly deploy fixes and being able to see, um, you know, analyze and access and remediate any of the vulnerabilities they have into the space and being able to do updates fairly quickly from a single point of view, whether it's sitting in a cloud or whether it's sitting on-prem.

Now, the probably the best way is to bring in a little bit of analogy in terms of why did is low cost carrier so successful? Where did they come from? How did they manage to do this?

How did they break this into the, the, the, the market? Um, similar probably to how we see our client providers doing something that very similar a few years ago, you know, 10 or 12 years ago. They've done also sort of a similar change within the market.

And one of the key things that they really achieved is through cost savings. And that is using a standard aircraft across the range, same parts, same maintenance crews, same engineering that they need. But also if we take it a little bit further, so that is sort of analogy to our operators, but we take a little bit further to their pilots.

The pilots need to be trained on the same aircraft. So again, some developers, again need to understand the same operating system, improve security inevitably day cabin crews know exactly what is the standards and features on this aircraft. They have one training module they have to go through to be able to stand this.

And this is why we're trying to bring this back to our operating system and why we're saying why do you need, want a single operating environment within your, um, landscape. So some of the common deployment challenges that we see a lot of customers having in terms of deploying it is that they have to deploy at scale. And it becomes, when it comes at that scale, it becomes more and more complex.

Do I manage a different Linux sitting in the cloud? Do I as to what I'm having on-prem? Do I use a different Linux again at my point of sales systems that's sitting at the edge that I'm having to use for my accounting systems and so forth?

So we're trying to find a way that we can comp, we can manage this, uh, very complex at scale environment consistently and be able to have a standard operating environment as they have standard aircraft, a standard operating system within our environments that the Linux operators know and love, the application developers know and love and everybody can be able to be very quickly in terms of provisioning and updating, managing this environment. But why a single operating environment? And it comes down to efficiency, and I'll show you in in next few slides where RA has done real work and, and we as red done real work with AWS to be able to very quickly be able to help customers, to provision new environments out there, to have standards, um, deliveries of the operating system to be able to simplify the, the, the support, the onboarding that we have.

Security is always such a key concept and it's not just being able to have the environment secure, but these patches and fixes need to be deployed very quickly. And if you have one single operating system that you need to, to manage and deploy this, this fixes to, whether it's sitting again in your on-premise environment and the different cloud environments in AWS very quickly, it can then be able to be compliant and be able to fix those security things. And then just having healthy patterns to sustain your automation.

And again, when we start looking at how do I manage automation the same way as I would manage, um, it within my standard operating system in on the premise, I can use the same automation methodology sitting in the cloud. So what do we mean with standard operating in terms of red enterprise Linux? Well, first of all, rail runs in almost anywhere you can think of that an operating system is required.

And in the most majority environment, it's the physical bare metal environment. It's our virtual machine sitting in on-premise environment. We can have a private cloud where we want to run this in, we do it in the public cloud, for example, in AWS we can even have an edge devices, like I said, point of sale systems.

But I've seen rail being used in wind turbines in the North Sea. How do I manage it? How do I actually have that capability of having all of this disparate systems under one pane of glass that I can actually see what's happening, understand what's the vulnerabilities, what patches needs to be applied, um, what is the analysis behind it?

Well, red Hat satellite is the tool that we do in all of this. So all I do is extend my radar hat, satellite manageability to the pro. The two of them are AWS environment and at the same time, then group them, whether that system is running on AWS or OnPrem, they can be part of the same group being able to manage them, complete lifecycle management behind it, being able to do updates and patches and so forth.

I also have Ansible automation platform. That's another tool on our arsenal. Again, the same automation tools that I'm using on, on-premise environments I can use for the same row environments sitting on my, uh, cloud environment.

It's the same Linux, the same applications can run the, the same playbooks can be executed in the different environments in the game single team that can manage both my cloud and my on-prem and my edge devices. And finally we have Red Air Insights, our SaaS solution that has the capability of being able to look at the cloud environments, including your on-prem environments. Again, but the beauty behind this, we have the advisory level behind within that.

So it will actually scan and look at the current environment. You have not the applications, not the data, just the operating system and advise you on how to fix certain things, how to improve security. WhatsApp fixes needs to be applied according to these things.

What is the best packages to deploy in within an example I have a little bit later I'll show you how to do those deployments into that environment. And that sort of creates a whole ecosystem around, well, for us in a cloud environment, on-prem environment, edge devices and so forth. And if we look at a bigger picture of where Red Hat Enterprise Linux really plays, I've spoken about the management side, I've spoken a little bit about the deployment, we'll go into that.

But we are really strengthening in, in terms of security and compliance, making sure that you can harden the system environment. Quite often customers, especially financial institutions, governments come to us and say, I need to have a hardened environment, CIS level two or Ansys Gap or P-C-I-D-S-S for the financial institutions. And again, we can do and help with the security compliance to be able to level up with those developers.

We still have a huge amount of developers out there that needs to develop and deploy in this environment. We have a broad ecosystem to be able to help developers. And again, that can be deployed within your AWS environments.

And then just in terms of the consistent performance you have, if you run rail on premise, you run rail on your, uh, uh, cloud environment, you know exactly what you need. You know exactly how to manage and migrated, and then migrations going back and forth. If you want to migrate from on-prem into the cloud, same operating system, same application, same management styles you need to do, if you need to do upgrades on-prem and in the cloud space, same upgrades apply to the same operating system.

So again, strengthen the reason why you want to have role in your cloud environment similar to what you have in your on-premise environment. Now, one of the key great features that we've announced with REL 10 coming, um, out of our uh, um, summit event last, uh, two weeks ago, is that we have worked very closely with our cloud providers to provide an optimized Red Hat enterprise Linux for each cloud provider and especially with AWS as well. So benefit from the seamless integration that's designed to work with that cloud provider environment, making sure that the tools are working, that the telemetry and the information you're getting from that.

And again, that those specific, um, environments for the cloud providers is being pushed into the marketplaces. And for the AWS environment, this is the red it will now find within the marketplace. Again, close integration with CloudWatch using the Open telemetry tools, making sure that the network performance is part of the Elastic network adapter and have it all the a WS two CLI tools within the operating system itself.

And those are available now from Marketplace specifically strengthened, applied, and made sure that they're optimized with AWS working very closely between Reddit and AWS to create you the best operating system that we can deliver. But also if you don't want to go with the marketplace, there's a different way of doing that as well. And again, we're trying to help our customers to make better decisions when they actually want to build these things.

And again, from the operating system point of view, our system operators, how do we do that? And within the red, a insights, even if you don't want to use the telemetry and you don't want to use the application tools, I suggest go and have a look at the tool, which is called Image Builder. And that is the ability to be able to create an AWS image that I deploy from scratch, but it's taking the latest packages, the latest details that I have, I can harden that.

I can add additional re repositories, additional applications, create the whole blueprint. This is not the final product. This is a blueprint.

It's almost like a recipe. You know, if you get a recipe for your good amisu putting, you want to dessert, but you only do that once you just before the guests arrive, but you don't want to steal all one out there. Similar with the blueprint in Image Builder, I tell it exactly what I want in that image, but it only gets deployed when I need it.

And then we'll pick the latest and greatest RPMs, the latest updates from the packages, the fixes, the security fixes, and we push that out into the environment. So lemme quick go through a few just, um, examples of this. This is, um, from the image builder.

com. I've re added all my details in terms of A-D-A-W-S, um, details. I've added some additional repositories and so forth in this instance.

Then I'm gonna select that I want to deploy this specific image to the AWS, but I can pick any of the others and I can also have on-premise environments like VMware and so forth. So the same package, the same blueprint I'm using here. I can push into other cloud providers, not just AWS and also to my VMware or my um, KBM environments.

One of the key things, of course, like I said, is that I can then pick the specific, um, uh, uh, AWS account I want to use that is getting immediately in there. It sort of defaults to the region that I typically associate with that, um, account. And then the next step is that I can then the, the decide if I want to do hardening.

Now, this is one of the things I really want to highlight with our users in AWS in the past we've seen some red out images being floating around that has hardening in that it doesn't come from red out. So please, I i I just wanna portion our users about that rather use this tool that we have with here, we have about 25 different levels of hardening you can pick in. There's specific to your PCI or whether it's in, um, for example, that I have the CSI level two or level one server.

And we will create the hardening right at the bolt time for you for that specific image and that you can then push into your AWS environment. The next step is where we do AWS package selection. I've just brought up there that I'm picking the AWS um, uh, CLI tools in there, you of course can add all the other tools that we typically put in the, uh, AWS environment for our marketplace images such as the telemetry and so forth as well.

And again, you can add in your own repositories into this tool as well with additional packages that you want to deploy into that specific image. And then finally, I can deploy this finished built image into my environment. The blueprint will be ready and I can redeploy this over and over again and again, each time I deploy, it'll build it from scratch with latest RPMs and latest fixes is in there.

I can select the size of the image I want the count of the image, I can add additional information, a few screens I didn't show, but I can add in my SSH keys into that. I can add in additional users time zones. There's a whole RAF of tools that I can use this for my deployment if I want to do my own deployments in AWS.

And again, this is where Reddi really comes into play, where we're trying to make your enterprise Linux deployments that you want to put into this environment as easy and as quickly as possible. Whether you want to bring your own licenses as simonus, as route, or you want to pick it from the marketplace, the offerings are both there and you can be able to pick and choose between the two of them. And then we've announced real 10, as I said, uh, a few weeks ago.

This is a really a significant release for us. And I just wanna highlight four key things that, um, we've brought into this specific release. The first one is image mode.

Um, this is a container technologies that is being brought into your, um, operating system so that the operating system, the virtual machine, the BareMetal deployment, the AWS deployment is behaving more like a container than previously it did with, uh, an operating system. And again, we can, you can, we can discuss this more in the summary section, the QA section afterwards, we have command line assistant, um, that is an AI engine that's within the command line itself. So if you forgot how to do a certain command, I certainly do because, um, you know, uh, we do forget things that we've done two years ago.

I can ask the command line assistant exactly how do I actually execute that command or even write me a little bash script and so forth. One of the key things that we are looking at is, um, when quantum computers come online, we are really worried about how quickly and possibly it's the ability to be able to decrypt our current encryption keys that we have if we get to quantum computing. So again, we're starting to apply a lot of post quantum encryption algorithms into the keys for L 10 to to, to be preemptive to that going forward.

And then finally, within the satellite I mentioned this is one tool that be able to manage both your on-prem and your and your, um, AWS environment altogether in one single environment. We're now bringing that advisor label that we have an insights into your satellite environment as well to be able to help you very quickly with those. And finally, back to you Simon.

Yeah, so thanks, thanks for that detail about RL as was called out. We can talk about the, um, new features of RL 10 in the summary session. We've got after the breakouts from Cloud Fridays, just to say we've quickly added some QRS here, so you'll be able to go if your interest is peaked and do some more research on the technology.

And of course, we've got upcoming sessions. We've got, um, sessions where you can pick, um, the session to talk about automation through and small automation platform on AWS or how you should maximize, um, spend through marketplace on AWS. Um, but all these resources are recorded and will be available for you to catch up on if you had to choose one, but you'd like to understand both.

And we do encourage you to watch all these sessions we think are very valuable for our customers. And that leaves us with just saying thank you. Thank you to my fantastic, um, counterpart here.

Jan, thank you very much for your session today. It's great to hear the detail you bring and thank you from us to you as customers for sitting through this session and joining us on Cloud Fridays. Simon, thanks so much for inviting me and thanks to everybody who joined us.

Hi everyone, and welcome to the six five Summit AI Unleashed. I'm Will Townsend. I manage the networking, telecommunications, and cybersecurity practices for more insights and strategy.

And today I'm joined by, uh, Marvel's VP of technology, mark Berle and VP of product marketing. Rishi showed for this cloud infrastructure spotlight on expanding the boundaries of custom silicon. And with that context set, gentlemen, let's jump in and, uh, mark, let's start with you.

Sure. And I've got a guest to this question, but, but I want to hear your perspective. You know, for decades, merchant Silicon has really dominated, you know, the data center.

And from your perspective, what's giving rise to custom and how is this different than the old ASIC business model Y? Yeah, I think it's really a couple of a couple of factors, right? First of all, uh, what we've seen over the last several years has been a pretty pronounced slowing of Moore's law.

Um, so unfortunately every technology generation, we don't get the doubling that we used to get. Unfortunately, uh, the data center doesn't care that Moore Law, Moore's law is slowing. Mm-hmm.

Uh, they need to find a way to, uh, increase performance every generation. And what that means is that, uh, there's a big change in emergence in things like chips to enable these customers to take advantage of these bigger, more complicated systems. And so, I mean, I would assume AI workloads are, are driving the need for that as well, right, mark?

Absolutely. And you know, the data center have really specific workloads that are really dedicated to specific functions, and they're different for each data center. AI is a big part of that.

Sure. And Rishi, any anything to add to that? Yeah, Exactly.

Adding to that stuff. Right. Uh, today's, uh, infrastructure deployment and the data center are more focused to a specific workload.

And, uh, and the service providers are basically constructing these data centers with that opex in mind. They wanna basically optimize their OP opex and give the maximum returns to their customers in terms of quality of service. And this basically fuels that custom silicon comes into play rather than the standard product coming into play and coming with the AI front as well.

AI is very power hungry, we know that. So there is heavy load of customization and a unique way of doing the things so that you optimize the power to get maximum returns on this investments and also your opex to make you really profitable. In fact, what we see that 25% of the compute silicon moving forward will be more towards the custom side of stuff.

Sure. And the project, what we are seeing on the custom XPOs are kind of in that direction of custom. And this basically trend has really become a norm in the industry, especially for the hyperscalers who are building their own data centers.

It really is. I mean, and, and the hyperscalers are doing their own silicon designs as well. Right.

So purpose-built Silicon is really sort of driving the train. And, and Rishi, mark, you know, mentioned chip, so I'd love for you to solve a debate. So, um, are CHIPS considered custom?

So there are two aspects of chip, I'll let Mark speak about it, but just at the high end, the whole idea of CHIP was done for two different reasons. One reason was to accelerate the execution time if there is a repeatable function within the organization or within the vendor who wants to put the footprint in all of the silicons. So it'll be more of a cut paste and copy to reduce their opex to doing it and enforcing the trademark on their silicon.

That's number one. The second one, uniqueness is also with these designs getting more complicated and the limitations of the die size or the radical size of the dye coming into play, uh, triplets becomes a more viable option where you can basically implant in it. And also if there are forward looking changes, it can be done very quickly without changing the parent mothership dye.

Okay. And Mark, anything to add there? Y yeah, and, and I want to kind of, um, circle back to what I was, uh, speaking about, um, for the first question, which was really about, again, Moore's loss slowing down, right?

We started to see Moore's loss slowing down in about seven nanometers, uh, moving, transitioning into the future technology nodes. And once we saw slowing down, as I mentioned, we couldn't sacrifice performance. So the amount of silicon that we needed to put into a package is, has been increasing and increasing every technology generation since then.

Mm-hmm. Um, now by the time we're in, you know, two nanometers with designs today, uh, we're looking at, you know, over a thousand square millimeters, 2000 square millimeters of silicon that really needs to be integrated together on a package. And the only way to do it is to break it into pieces using technologies like chip.

Mm-hmm. Um, so in a way, chips, uh, don't have to be custom, but a lot of these big designs, the only way to achieve that performance is to use multiple chips working together. Yeah, no, that makes, that makes for perfect sense.

So let's shift the conversation to memory. And you know, I, I spent a lot of time with the Marvell team and, and I know this is a big focus for, for Marvell and historically memory has been, you know, based on standards for obvious reasons for, for scale, you know, supply chain and costs and that sort of thing. But I'm beginning to see the rise of custom memory, like custom, you know, high bandwidth memory and that sort of thing.

And, and so Mark, from your perspective, why are we seeing this? Is the answer AI again, or is there more to it than that? Well, AI is certainly the major driver to this, but, but again, it kind of goes to play with the, the incredible need to get more and more performance.

What happens with customizing HBM, especially with Marvel's custom HBM solutions, is that we can pull a lot of the content that was just interfacing to those memories off of the main die, off of the main die for an accelerator. And that opens up a lot more area for, for our customers to really get the performance, you know, the, the number of tariff flops that they need in the design. Mm-hmm.

So that makes, that makes a big difference. Like, for example, we can clear up like 25% of the dice space that would've otherwise been dedicated to talking just to HBM memories themselves. And we still get even better performance of the custom HBM with about 70% less power spent on the interface to those memories themselves.

And, and, uh, I'm sure as, as you're aware, right, power has become really, really critical for a lot of these applications. And Mark, you sort of touched on this when you were talking about Moore's Law and, you know, hitting the wall there. I mean, custom's been long associated with XPS and CPUs, right?

And I mean, you look what, you know, others are doing with, you know, GPU technology, AMDs and the NVIDIAs of the world and, and adjacent technologies like, you know, and packages like memory, but it is it expanding even beyond that base as well. So we, we spoke about how customizing HBM is really, really important when we're integrating high capacity memory on the package. Um, one thing that's equally important is actually, uh, the sram, the embedded SRAM that's, uh, actually built within the accelerator or switch or, or whatever, uh, kind of customized data center device you're building.

Um, we actually, uh, are building, um, very specific highly customized memory solutions, uh, for our, for our customers, um, that are really cranking out as much bandwidth as we can per square millimeter of space. And in fact, our memory solutions are providing about 17 times the bandwidth that you can get from an off the shelf memory today. Uh, so it's a, it's a really big deal for us.

It's a really big deal for our customers, and it's kind of existential to continuing to grow performance in AI Beyond that, um, not only is that high bandwidth memory, the, the embedded esra memory incredibly important for these devices, but also high capacity, uh, memory is a huge need in the data center as well. And we address that actually with a, a set of, uh, product families that are optimized to, uh, allow CXL connectivity for, um, memory acceleration and for memory pooling or capacity enabling like 12 terabytes of memory accessible through this one, uh, highly op, a highly, uh, sized, optimized, uh, chip. So it's, it's a pretty, it's a pretty special offering kind of stacking all the different types of memory that, that are needed for the data center.

So not only, you know, do we customize accelerators, nicks, components like that, but we're even seeing customization, um, moving into the switch arena. And, uh, that's actually what Rishi is an expert on. So, uh, I'd really like to get his perspective on that.

Yeah, Rishi would love to hear that. Yeah. So on the, on the switching side, right, it's basically what it's switching all about.

It's basically routing large number of packets into different segments. So there is a lookup table associated with the switch, and there is also a huge centralized buffer scheme inside the switching. And as we increase erratic side erratics of the switch, that means more fan outs of the switch.

That means indirectly we are telling that there are more individual ports, which are talking to the switch, more unique Mac, Mac addresses. And plus there is also a cocktail of things going on between the layer two and layer three in terms of routing interfaces. Mm-hmm.

And this all has to be all encapsulated into a large buffer size and managed accurately to do the right pipelining for the switching, uh, uh, to happen with a very low latency. So memory inside the device or SRAM in the device. And the overall, I would say the management of those, uh, memory modules become very important for us in terms of fan out, in terms of also data manipulation, which is happening inside, and also at what rate the, uh, bandwidth rate are we operating at.

So overall, this whole infrastructure and the SRAM implementation on the switching side and custom switch side becomes very, very critical. In fact, I'm going to speak more about it when we talk about the different criteria of switching within the switch and this memory, you have different bifurcations happening because of ai, which is scale up and scale out. And also within this bifurcation of, especially on scale up, there is much more requirement for memory because scale up is something where yearly you want to utilize all of your GPU optimization, the highest opex what they're spending is on the GPUs.

And the last thing you want is to be underutilized. And the under utilization of this GPUs are coming because of the memory. So some GPUs will be starving from memory, some GPUs will be kind of underutilized for their memory, uh, usage, what they have.

And the switch fabric is the one which kind of capsulates all of them gets in into a unified structure and scale up and provides a unique value proposition in terms of custom, where you can make sure that your investments on your infrastructure, especially on the GPU side, is kind of amortized across all the nodes. Yeah. And you can actually, you know, fine tune the performance as well, right.

With Absolutely. And you know, the, another trend that I'm seeing is the integration of DPU and into top rack switches as well, um, to embed network and security services. Marvell has a, uh, has a DP line, I've, I've written about that, um, you know, over, you know, over the, you know, the last few years as well.

So there's, there's a lot of innovation that's coming, you know, within the switching segment. Mark, any, any other perspective on, on the whole Smart Switch evolution? Yeah, and, um, and this really speaks to the topic for discussion today, right?

Customization is, is really everywhere. And I think we're gonna continue to see, um, especially with, with innovative, uh, NIC solutions, we're gonna continue to see what was a very, um, let's say, uh, standard and well-known, um, architecture for, for switching, uh, really evolving to, to be extraordinarily unique, um, and, and very specific to, uh, to different customers and their, their implementation. So, uh, we're, I, I think we're going to see exactly the, the trends that you alluded to, uh Sure.

In a continuing, you know, in more and more, uh, than we have in the past. Yeah. Well, as we round out our conversation, that's a great segue to, uh, the final question I have for both of you and Rishi, I'll start with you.

Um, what's next in the evolution of custom from your perspective? A good question, actually. If you look from the whole phenomenon of custom, and I would start, if you look at, uh, data center rack, all the big iron components sitting inside the rack, we have already seen, uh, customization happening on the CPU and the XPU side of the components sitting inside the rack.

And now we are seeing, as I mentioned, the bifurcation of different classes of switch. There is a traditional cloud switch, and then there is an AI switch within the AI switch. There's further bifurcation within the scale up and scale out switch.

So it'll be of no surprise that the next frontier of customization will be on the switching side where people will be doing custom switches. And also this goes in hand to hand with the custom XP dpu and the next what we are seeing it, we are also seeing, uh, a lot of push for the, of, for the photonic side in terms of the photonic engine or the optical engine being integrated inside the silicon, which is the CPO. Mm-hmm.

And of other frontier will be, there are different technologies, and definitely customization would be playing a very crucial factor in CPOs because CPO engine by itself will be defined based on your usage, your scalability, what length you want to drive, what applications you want to drive. And also in, in order to get these things done, it goes a further level into platform where you need to look into the customization at the platform level. Whether, what reach are you using, are you using copper or optics?

What thermal solutions are you using on your design? Either it is liquid cool, or it is just air. Cool.

All of these will be factored into the custom silicon part of it. It'll not be just a silicon engineering, which has been customized for application usage, but it'll be a platform and system which will influence the customization to happen. So as this rag goes through a transformation phase from CPUs to XPOs, and we have already seen the phase happening for the nick, we have done a custom nick for a lot of our hyperscalers, which is in the public domain, and now it's basically the next evolution will be on the connectivity and on the flip side, Yeah.

Networking is sexy again. Right. Mark, what, what do you see in your crystal ball with the evolution of custom?

Yeah, absolutely. And, and, uh, really building off of what Rishi shared with us, um, I think we're going to see, um, more and more complexity integrated with these devices, um, as, and as we essentially define this platform that allows, that allows the data center to kind, uh, achieve their goals from a connectivity point of view. So we're gonna see more and more integration, uh, things like co-pack, copper, co-pack optics, uh, really driving really complex package technology.

Mm-hmm. Uh, we just announced recently, um, you know, a, a novel multi interposer two and a half d uh, package integration technology platform that allows our customers to really scale up the amount of content and also to, uh, simplify two and a half d integration by removing silicon from the equation. Um, so these kind of innovations and investments in the platform are really going to continue to grow this trend of system in a package, um mm-hmm.

Which is really, really key for, for the data center to really achieve its performance goals. Yeah, and I'm, I'm glad you mentioned co-pack optics. I mean, that's a pretty hot topic right now, right?

I mean, the efficiencies there, the performance improvement, the, the power management that comes along with that, uh, that's an area that I'm beginning to dig into and, and hope to publish some, um, you know, some insights on in the near future. So, but gentlemen, thank you for your time. It's been a very, very compelling conversation.

I want to thank all of our audience for joining us for this cloud infrastructure spotlight at the six five Summit. Stay connected with us on social media. com/summit, and there are more insights to come.

So stay tuned. Welcome back to the six five Summit, and we are talking about one of my favorite topics. Okay.

It's my favorite topic, and that's semiconductors. And we're joined here with submit, uh, from Micron. Uh, you know, HBM storage memory is becoming absolutely paramount to be able, uh, to fully execute this build out, uh, in ai.

And I really appreciate you joining the show. Thank You for having me, pat. Really excited to be here.

Isn't it amazing? You know, I, I, I celebrated my 35th year in tech this month, and, uh, memory and storage, man, it's like, you know, you better hang on. It's gonna move, right?

Roller coasters going up, rollercoasters going down, but man, it is up, up, up into the right. And when I started hearing about, uh, things like, you know, HBM as the catalyst, uh, of the industry, and then all the stuff on DDR five, I, I'm getting way ahead of myself. So, uh, I, I'm here to ask you questions here, but thank you for, uh, thank you for joining us here.

So, uh, first question is, is how are you rethinking the role of memory, uh, and storage, uh, in enabling all these next generation AI architectures? I mean, regardless of, you know, hyperscaler data center, tier two, neo cloud client computing, the, the car, I mean, a AI is literally, um, um, everywhere. Absolutely.

Uh, AI is such an exciting secular growth trajectory for us and for the whole industry. And, uh, we have, um, just reported earnings yesterday. Um, you have seen a lot of the growth that AI is driving for us.

And I think if we take a step back and look at the overall momentum that AI has gained over the last, uh, couple of years since the advent of chat GPT and generative ai, um, we see it as four important exponential trends that have come together. The first one is the radical improvements in compute architecture and compute capability. These have been driven by GPUs, uh, which have really enabled, uh, AI based hardware to get to the next level of capability.

And very much hand in hand with the compute capability. Is the memory subsystem capability required to feed the processor, all of that insatiable amount of data. So, um, really the exponential trends on compute capabilities is the first important one.

The second one is the dramatic improvement in software capability. And of course, generative AI itself was a big leap because, you know, you look at the last couple of decades, uh, AI field has been making progress for many, many years, but this was a huge step function, um, with generative ai. And now of course, a lot of new ideas coming to the forefront to make generative AI even better.

On the third one is the massive amounts of data that we all have access to, and how the models can be trained on immense volumes of data. You've heard about trillions of parameters in these models, um, from billions to hundreds of billions to now trillions of parameters. So a huge amount of data.

And these models through the hardware and software capability are able to extract a lot of that insight and knowledge and intelligence out of this. And the fourth one is the ubiquitous connectivity between devices that then generates even more data and feeds on this trend. So those four exponentials coming together, and it starts off in the data center, um, we have huge growth in the data center that is going on.

Uh, that has been what has underpinned our results that we just reported yesterday. Yeah, with 50% sequential growth in HBM alone and high bandwidth memory is now more than a $6 billion annualized run rate of revenue for us, which is, uh, which is great. Um, and so, you know, when we think about this whole AI revolution, if you think about GPUs as the brains of ai, then you have memory as the heart of AI because, you know, just like the heart pumps blood into all parts of the body and enables the brain to do its thinking, uh, we really think that all of this data that these processes have to process that memory processor bandwidth, uh, oftentimes, especially in inferencing, becomes the bottleneck of system level performance.

And so we are doing a lot of work with our customers on how to make these memory subsystems a lot more capable. And once it, you know, grows, it is not going to just stay at the data center, it's going to go to the edge, it's going to go into automobile's, uh, PC, smartphone. So really exciting time.

And we are just at the start of that big wave of AI growth for the next, uh, at least a decade, if not two. Yeah, I, I agree with the, the decade. It is exciting.

I was one of a hundred people that was invited to the original, uh, announcement with Satya and Sam Altman. And, um, I was sitting in, I was sitting in the audience having a lot of thoughts. One of them was about architecture, uh, and data.

And that, you know, it's a, a well-known understanding now that, you know, there is indeed a memory bandwidth bottleneck, uh, out there. It is a challenge, uh, for, for ai. I mean, it's hard enough to have, uh, an application address, uh, a ton of memory and, and a large model.

But now, you know, we're, we're into this idea of reasoning engines where even inference is, is difficult. Uh, I'm, I'm curious, uh, you invest a tremendous amount of money into r and d, um, and you know, you've got multi-year roadmap. Like how are you looking at, um, exploring, you know, I've seen you talk about new memory compute architectures to address this, This challenge.

Yeah. You bring up a really important point. This memory bandwidth bottleneck, uh, is an important one to, I wouldn't say solve because it's not easy to have the memory bandwidth grow at the same rate at which, you know, number of cores and GPU performance grows.

Um, but at least to improve on that in as meaningful a way as possible. And so, uh, if you step back and think about the bigger problem, you have the memory bandwidth issue, and you have a power consumption challenge in the data center in aggregate, right? So there are a few ways in which you can solve the memory bandwidth challenge, or at least improve it, if not solve it.

But they may not be the most efficient ways, or they may violate some other challenges or boundary conditions related to power consumption, as an example. Um, so, uh, our focus in all of the innovation that we have been driving is how do we make power consumption a big focus for the company in terms of, uh, improving it from one product generation to the other, as well as improving it versus the competitive bar, right? So that has been a huge, uh, focus for us.

Another big focus has been, uh, really having leadership on the process technology side, because every new node of process technology gets us better performance and more power efficiency. And the idea behind both of these is, as our products become more power efficient, they enable the GPUs to do more. And you can think of it in two ways.

You can either have a data center consume less power than it otherwise would with these power efficient capabilities or within a given power envelope for a data center, you can run more compute than you would otherwise be able to because of all of these power capabilities. So in thinking of the memory bandwidth issue, we are coming at this is this problem in from many different angles. Power is one very important one because, uh, if you take the example of HBM three E microns, HBM three E has 30% lower power consumption than the next best competitor.

And this is in an industry where one, two, 3% power consumption differences are very typical, um, between competing products from different companies. So 30% is like an order of magnitude better power consumption, um, versus what is typical between competing products. And what this does is it enables the processor to run at a higher performance within the same power envelope.

So if we improve memory bandwidth from that angle. Now, the other thing that we do to reduce overall power consumption so that the aggregate memory bandwidth, aggregate memory capacity can be increased within the same power envelope is to focus on bringing LP dram, which is low power DRAM into the data center. Now, low power DRAM has been thus far only relegated to, uh, PCs like laptops and smartphones where battery consumption and battery powers, um, and how long a battery lasts are critical factors in those devices.

So they end up using low power DRAM in those products. So Micron had the idea of taking low power DRAM into the data center, and we have to solve a lot of technical challenges to do that. Most importantly, RAs type of challenges, reliability, availability, and serviceability, because the DDR R five based products have a very different RAs profile than LP dram.

And so we pioneered the use of LPD RAM in the data center. It is now shipping in volume with one of the largest, uh, customers in the world for AI products, um, and AI memory. And it has become a very significant business for us.

We have just mentioned that, uh, LP DRAM and um, and high capacity DIMS in the data center have together become a multi-billion dollar business for us. Uh, so this is another big way where as we take LP DRAM into the data center, processors can connect to more dram Yeah. And have more aggregate bandwidth in helping solve some of this memory bandwidth challenge.

Because within that same constrained power envelope, you can just have a lot more dram, um, in the system. And the last couple of ways I will mention is, um, four HBM, um, our HBM roadmap on HBM four E, we have just sampled HBM four. The next generation after that is four E for hbm, four E, we are doing custom based eye design in partnership with our largest customers.

This custom based eye design enables our customers to move some of the blocks of logic from their, uh, processor, GPU or ASIC onto the base D. And that enables a much better, uh, power efficiency performance as well as, um, memory bandwidth in the aggregate system. Uh, and by aggregate system, I mean, whether you're connecting just to the HBM or you're connecting to LPD RAM around it.

So the overall system design just becomes higher bandwidth, higher performance, um, much more capable. And this is a tremendous effort, uh, done in conjunction in partnership with our customers. We are very excited about it, and this is another important vector.

And the last one I will mention is, um, we are doing a lot of innovation on our roadmap related to new architectures, new approaches. One example is bringing processing closer to memory. So think of what is called PIM processing in memory.

Uh, that is one approach to do this. There are other approaches that we are pursuing as well. And in all of these approaches, there is varying level of impact to how the software stack in the system, uh, sees the memory, and does it need, does the new architecture need changes to the software stack or not?

And, and depending on whether the software stack needs to make changes to see this new architecture, use this new architecture, obviously it creates a new level of friction in the adoption of some of these new architecture. So we are going through the pros and cons of many different approaches on how to solve and improve the memory bandwidth issue. We are working closely with customers, um, we are running different, um, uh, different test cases with them.

And it's a very exciting, uh, time to be in memory because there's so much opportunity for differentiation, uh, which is what we love. Yeah, it is a great time to be, uh, in memory. So as, as I step back and reflect on what you just said, just the amount of innovation is just, is mind blowing.

Uh, a lot of people I know get fixated on the compute side, but quite frankly, the compute can't do anything, uh, without the memory here. And on the data side, um, as well, the, the ability to store, I mean, we're creating, uh, more data in a year than, uh, existed, uh, or has ever been created, uh, uh, before it. Uh, and I, I think the way you're approaching it by first of all, um, L-P-D-D-R in, um, as a, as a way to reduce power in another area that can lower the overall, uh, power footprint and increase the density, uh, while reducing the power with HBM three E, and then in the future, uh, with four E by the way, I, I have done some research on custom, uh, HBM mm-hmm.

Uh, four. Um, and I even, uh, did a video with one of your partners at, with Marvell, um, as well. So I, I find that, uh, fascinating, uh, in-memory processing is definitely on our, on our, um, futures look, uh, here.

And what's funny is, I, I knew I, I think we're gonna get there. Um, and the whole power I, I think is the biggest driver, which is, uh, is is pretty cool to see. But, but let me move on here.

So I, I talked a little bit about, you know, right now the fixation seems to be an AI on the hyperscalers, and I, I get it, right, uh, the CapEx that's being invested. Uh, you're a huge beneficiary, uh, of that. But the reality is, and we've seen this time and time again historically, is, is that, um, at least our thesis is right, this moves from hyperscalers tier two CSPs to enterprise ai, uh, to the industrial edge.

Uh, and then finally, and by the way, this is in no, uh, order of when it's gonna hit, uh, devices like PCs, uh, and devices like, uh, like smartphones. I'm curious, how are you enabling, or let me use the word, democratizing AI for these different, uh, use cases and platforms? Yeah, we think about, um, the concept that you just brought up about democratizing AI for various categories of customers.

And we also think about how we democratize AI for the masses, right? Meaning, yes, adding more people in society have access to more of these AI capabilities. So let me sort of touch on both of those, because both are important in different ways.

So if you think about the, um, data center first, and you're right, I mean, right now a lot of the CapEx, uh, driven out of the largest hyperscalers in the world, they're building out all of these AI factories. And these AI factories have huge amounts of compute capability. And of course, you know, GPUs, um, asics, uh, for processors, um, massive amounts of memory.

And of course, these, these AI servers also have huge amounts of, uh, SSD data center SSDs. And that's a big play for us as well. We are very proud of our trajectory of getting to record share quarter after quarter in that as well.

But as we think about expanding, you know, all of this, uh, you are seeing so much of an embrace of AI from sovereign countries wanting to implement AI in these data centers in their own country for their own societies, and have the AI be trained on their own localized data sets. And, um, this is important to a lot of countries for their own, uh, country's growth purposes. And so lot of these companies are coming up around the world, which are focused on those markets, and that's going to be, uh, a growth market for many years to come.

Uh, we are also seeing OEMs, um, that supply to some of these hyperscalers, these AI servers, but are also starting to supply these AI servers to, um, captive, uh, smaller enterprise installations for enterprises in the Fortune 1000, for example. And, uh, there is some data that companies are happy to locate into the cloud. There is some data that for security and privacy reasons, has to stay, um, within the four walls of the company.

And then these private clouds or on-prem type of installations become important and training of the data sets that are company specific with very confidential, um, highly secure requirements also create, um, that opportunity for that part of the market to grow quite a bit. So within the, uh, data center, uh, there are various fragmentation of, um, pockets of growth that are going to continue to, uh, become much bigger in the future. And I think, as you pointed out, this is starting out in the data center, but it's going to proliferate from there.

The industrial opportunity is just massive. Initially, it's going to be more AI going, getting infused in a lot of your existing industrial devices, but you can very well see huge amounts of automation being enabled by ai. Huge amounts of growth coming from a new field in robotics.

Robotics has been around for a while. Automation through robotics has been around for some time, but it is getting new wings of growth because the capability of these robotics subsystems and even, you know, humanoid robots in the future are getting accelerated with the use of AI in, you know, enabling that acceleration. And then when you combine the conversational capability that generative AI has created to, you know, create knowledge and insights and intelligence outta massive quantities of data, and you combine that with, you know, robotics in the humanoid form, now you can get extraordinarily capable, uh, assistance of all kinds, whether you're doing, you know, dangerous work or you're doing, you know, repetitive chores.

Um, there is a huge opportunity for robotics to grow over the next 20 years, um, in a very, very big way. And then of course, you know, you look at automobiles, you are seeing the growth of autonomous vehicles with Waymo having completed, you know, so many tens of thousands of rides. Um, and now Tesla launching its service.

And of course, these are early days. There'll be ups and downs in this, but the technology's evolving super rapidly, uh, with the help of ai. And I have no doubt that this trend is going to gain momentum over the next five, 10 years as well.

So there are a lot of these structural trends, and as you mentioned, PCs and smartphones are going to have a lot of new compelling applications. I think that underlying capability exists today already to create these applications. Now it's a matter of, you know, perfecting them, ensuring the right level of privacy is maintained so people don't freak out when, you know, your own personal data gets accessed in ways that you didn't envision or, or didn't, you know, heaven forbid agree to.

So all of those things have to be taken care of. So it's taking a little bit longer, but, uh, there are applications of this that, um, don't have to wait very long. For example, you know, real time translation, being able to speak to people, uh, with different, uh, linguistic, um, uh, languages, different languages in a seamless real time way using, you know, just your phone, uh, or your PC on a zoom and, uh, making it just, uh, a very, very simple experience versus a frictionless experience versus what it has been in the past.

These are just very simple examples. Yeah. And then, you know, when you think about, um, democratization for the masses, right?

Think about access to all human expertise on these cloud platforms with the very low cost per token. You know, as we keep improving the hardware platforms, the software, the cost per token keeps going down. And now you can have very broad access for the masses of all kinds of expertise.

I mean, you want to draft a legal letter, Chad, GPT can do that for you. You have a question about some odd symptoms you're having, and you want to know what it could be before you wait two weeks to see the doctor, or two months to see a doctor. You can get a lot of input from, from these generative AI models, um, and think about the evolution and revolution and education that can happen.

Yeah. Um, so much tailored, um, education modules can be created for students of different, uh, caliber within a class. And it can be a huge aid for teachers that can totally revolutionize education.

So I think in any field, you look at the sky is the limit in terms of not just productivity, but bringing the benefit of AI to the masses, um, as well. So really exciting times ahead. Wow, you said a lot there.

Very meaningful things. Uh, you know, even as recent or at Davos, I mean the, the biggest tech discussion was an air gap sovereign AI cloud, right? And now we see what popped up in the Middle East with, uh, with all the deals that, uh, that, that were cut there exasperated by tariffs, right?

Yes. So that discussion, um, you know, I'm, I'm convinced that, uh, that, that there is 100% something, something there. Everybody wants their own AI cloud, uh, and they want it to want their own.

Exactly. Uh, on the edge. I mean, what some people forget is, if you look seven, eight years ago when you know the edge was hot, um, now we have between a hundred and a thousand x, um, AI performance per watt, and you're just gonna be able to do, uh, a whole lot more.

Uh, uh, o old, old time robotics used to spend most of your expense and setting up the robots because they had to be within, um, millimeters and sometimes microns of perfection. Now, with object recognition, it acts more like a human where it adjusts to a, a, a changing, uh, environment. And I love what you said about democratizing, uh, in, in societal, uh, I know a lot of people like to be, you know, really down.

I know what we've done in the past 50 years, we've almost eliminated poverty, okay? Yes. Haven't completely.

And imagine if we could take that to education and healthcare would be absolutely game changing, uh, as, as society, I'm, I'm really glad, uh, uh, you brought that up, Sumit. This has been a great conversation. I, I like camp out here for an hour.

Um, but, uh, uh, I wanna wrap up with a question here. So, uh, my research firm tracks probably a hundred different semiconductor companies, uh, and we, we track, you know, three to four memory, uh, and storage companies. And I'm curious, how are you different when it comes to ai, I would say in aggregate in the big picture, but also across your competitors?

Yeah, it's a great question. Uh, pat, thanks for, um, having us discuss this because, you know, I'm very passionate about Microns place in the ecosystem. We are very proud to be the only US headquartered memory company of scale in the world.

Uh, and we are the only manufacturers of memory, um, in the us. So today, only 2% of the world's DRAM gets manufactured in the US and that too is done by Micron in our Virginia Manassas facility. Um, but we are on a huge project, um, that I, I know, you know, uh, we announced a $200 billion investment for r and d, uh, and manufacturing in the US over the next 20 plus years.

Uh, 50 billion in r and d, one 50 billion, uh, in manufacturing. And this includes, you know, building two fabs in Idaho, in Boise, uh, following that up with up to four fabs in upstate New York and modernizing and bringing one alpha DRAM technology to our Virginia fab, uh, as well as doing advanced packaging. And we didn't get a chance to speak a lot of our packaging, but packaging is a huge, huge deal, as you know, and you have spoken about that in the past, in your, um, uh, videos.

And so this is a, a hugely important thing for the US in terms of, uh, having packaging, leading edge packaging technology, uh, investments from us. And so I think, um, that is one huge differentiator in terms of, um, our US headquarters and our US presence and our US investments that we are making on behalf of customers. And our customers are very excited about that.

Um, so we really look forward to that. I think the other aspect is, uh, you have known Micron for many years. Um, our customers and people in the ecosystem have known Micron for decades.

Um, uh, we have been around for a long time, but this is not the same micron that people have known for for decades. You know, uh, this company today, uh, micron is better positioned competitively than at any time in our history. And let me just give you a couple of data points.

In DRAM technology, leading edge DRAM technology, four generations in a row, micron has led the entire world in time to market in coming out with the latest node of DRAM technology, and not by a small amount sometimes as much as a year, right? Right. So these are like huge, um, improvements in our technological capability, three nodes in a row on the nan side, leading the world.

Uh, our product portfolio has never been stronger. And I think the story, when the story of Micron is written in the last five years or so, our, our product portfolio has taken giant leaps and we are so proud of the work the Micron team has done and innovated with our customers because we have so many industry firsts and industry best type of products in our portfolio today. And I gave you two examples of that with HPM three E with LPD ram, um, these innovations are very meaningful to our customers.

They're not small deltas to our competitive baseline. They're complete game changers. And, uh, also, you know, when we look at, uh, data Center SSDs, you know, we have been on a very successful trajectory of improving our share in some of these most complex products, uh, in the industry.

And I'll just end there that, you know, because of the strength that I, uh, we have been gaining share in all of the high profit pools of the industry, uh, across the board. So, really exciting time. And as we discussed, and as you have mentioned a lot in your videos, uh, AI is just a phenomenal, um, multi-year, decade or two long growth opportunity.

And we are so excited to be able to go innovate with our customers, uh, for the benefit of, um, AI for all of the society. Yeah. On the differentiation side, I, I spent a lot of time with, uh, Jeremy, uh, on, on, on your side.

Yes. Past couple years ago. And the amount of firsts and the time to market advantage, uh, were, were evident.

And it's great to see, uh, you hitting on all cylinders on, on HBM. And I know the world is more than HBM, but, you know, striking, uh, while the iron's hot, uh, the whole DDR five transition is, is exciting. And, um, I think it's, you know, meaningful, particularly when it comes to performance server consolidation, uh, out there, uh, in the data center.

So, Sumit, I, I really wanna appreciate you, uh, coming on and, and taking us, uh, through that. It was great to see, great to see you again, and I hope we can do it again sometime. Thank you, pat, and really appreciate your time.

And, uh, as always, I've enjoyed our discussion and I look forward to the next one. Definitely. So that wraps up our semiconductor spotlight session here with Micron.

And whether it's, uh, the core Hyperscaler data center to the edge, uh, I hope you're all convinced out there that memory is fast becoming, uh, the most important element when it comes to, uh, ai. And Mike Ryan has a lot of firsts, uh, in it. It was great to see, uh, how they, they crush their earnings, uh, recently.

And I know earnings aren't everything, but it is a, a one very important metric that, uh, that shows that customers are buying a lot of memory from Micron and also they can fill their coffers, uh, to build out facilities and, and invest into r and d in the future. Hit that subscribe button. Thanks for being part of the community.

Take care.