Quantum Computing Breakthrough – Katie Pizzolato, IBM

This is Techstrong tv. Hey everyone. Welcome back here to techstrong tv.

I'm really excited to talk about some exciting news in the realm of Quantum today. Joining me is Katie Pido. Katie is the Director of Quantum Theory and Computational Science at ibm, and I hope I got all that right.

Katie, welcome to techstrong tv. Hey, thanks for having me. All right.

So Katie, look, it's, uh, quite a title, right? Quantum Theory and Computational Science. It's right at Rocket Science level.

But tell us, give us a little bit of your background, a little bit of your story, how you came to be the director here of this. Yeah, so, um, I, I am an, you know, an engineer by trade. Uh, I've been with I B M for a while.

I've, uh, been in different parts of the business, um, from, you know, chip packaging and other parts of growth. You know, growth tech, sector growth, parts of I B M, um, leading strong teams, uh, developing strategy and executing on it. Uh, so I joined the Quantum team about five, well, almost six years ago now.

Uh, and it's been quite, quite the ride. It really has, you know, five, six years ago, pre covid. I remember the, I, I guess it was an IBM think event, and they had this, I don't know if it was functional or just a model of a quantum computer.

It looked like a very fancy cappuccino machine. Machine. It was all copper and everything.

I, you probably know which one I'm talking about. Yeah. Chandelier.

Yes. It was a bit of a, and I was like, wow. You know, the geek in me and the gadget, boy, I was, I couldn't, I was mesmerized, right?

And, um, and I couldn't wait for it to get real. And, but it's getting real. It's, you know, yeah, I'm sure in your five or six years you've seen tremendous progress, Tremendous progress.

Uh, and I think that it's just, uh, there's so much progress. Cause there's so many people, you know, there's a community effort, there's obviously a lot of industry involved. Startup ecosystem is really strong.

Uh, it's a really exciting tech that we think is gonna do revolutionary things, and we're excited to see it continue to scale. Absolutely. So look for those in, and our audience is a very technical audience.

They, they know IBM obviously, and the IBM's long history of really kinda leading cutting edge technologies, making huge breakthroughs, whether it was on CPU u and hardware as well as software. But it seems like maybe we have an announcement you could share with us, or some news about a new breakthrough from ibm. Yeah, absolutely.

So what we know that there's a promise, uh, you know, the promise of Quantum is that it's going to be good at things, uh, good at very specific things. There's things that classical computers will always continue to be good at, and they will continue to get better, right? If you, if you asked in the eighties, if classical computers were mature, what would your answer have been?

Right? It was doing things that we thought was pretty cool, and now, yes, as we sit here today, it continues to do things that are even cooler, right? More, more than we thought then.

And it's gonna continue to evolve and continue to do really great things. So we see a very similar trajectory with Quantum. And as we have been able to get better and better hardware, we've been able to translate that into computational performance.

So again, you know, history of classical, we made better chips and devices, and we did things with those chips and devices. So we're looking, we're seeing the exact same kind of trajectory in quantum. So in a recent paper that i b m published with our collaborators at Berkeley, we're really looking for, I would say two things.

Like, one is how do we understand when quantum can approach problems that are beyond the, definitely beyond the exact classical methods, and start to challenge the approximate methods, and how do we verify those things? So, um, you know, if we're in this approximate regime, how do we know if we're beyond exact? And in an approximate, how do we know that the classical, that the quantum computer is performing with or beyond the classical?

So this, this, this, this project and this demonstrations experiment was really geared toward that. What can we simulate? How do we look at comparisons and how do we understand where we start to find those limitations and those breakdowns of those classical methods?

So that's exactly what we did here. We found that, um, the, the quantum computer performed accurately with the classical computer in these exact verification regimes. And then as we left the exact and entered the approximate, we found areas where, uh, the classical d the classical, um, methodologies started to break down, and the quantum computers continued to perform very well.

What is really exciting about this, though, is that the scale of which we were able to do this, so the, the, the, the experiment was performed on 127 qubits at a depth of 60. And you can just think about that as a very large, like a very large circuit and a, a very large space that, again, is beyond this exact kind of verifi regime, but it's a testament to how far the hardware has progressed. Um, but we're getting accurate results at that scale, which is, we've never, you know, getting accurate results of that scale is really unheard of.

And I think, you know, something that the team was really excited about, then we, you know, obviously started looking at how do we compare to these approximate results and these leading methods, but it's not gonna be, we never, we never wanted this to be like, this is a moment in time. We beat all the methods and we're done. This was a, you know, a method that we were looking at.

How can we, how can we actually devise the experiment in a way that we can compare to these methodologies? And how do we continue to explore, like these regimes where classical methods break down? Because if you can simulate it efficiently, classically, there is no quantum advantage.

You're really looking for quantum to be able to simulate things efficiently in a way that classical can't. Uh, so that's what's really exciting. We were able to devise a very large scale exer experiment, um, showing accurate results at a huge scale on quantum devices, uh, an unprecedented scale on quantum devices.

And then we were also able to look at how classic some leading classical classical methods started to break down, and quantum devices were still able to simulate in this regime. And what, what we really want from this is, do we want these quantum computers to be used as a tool? Like how do we find these circuits that we, that can start to, you know, challenge the, the approximation methods of classical, and then how do we find great things to do with them that is gonna be this path?

We've gotta map hard circuits to quantum devices, and then we gotta find really interesting things to do with them. And we are already seeing, uh, a lot of interest in trying to look at how we can simulate these things differently. And maybe, maybe quantum has to, you know, there's gonna be a back and forth with quantum classical.

This is not gonna be, there is no moment in time that cla you know, quantum is gonna say, we won, we're done. Uh, we are gonna get better. Classical's gonna continue to get better, and it's gonna be a really interesting back and forth to explore these approximate regimes that with new tools, Look, it sounds like hybrid to me, right?

I, I, I don't, from what little I know, and I don't profess to be a quantum expert, I think what you said, and I want our audience to take away this, if they take nothing else away, there are some things that Quantum is a game changer for. Yes. Right?

But there are some things where, quite frankly, our, as you call it, classical com computers are more than adequate and will continue to be more than adequate going forward. And if there is going to be a, uh, a greater cost of, of doing, if you have one task, and it can be performed by either classical or quantum, and the quantum costs more to do it well, obviously, right? Use the classical, use the right tool for the job.

I think what we need to understand as an, as a industry and, and as you know, society in general is what are the, what are the, the use cases for quantum, right? Where no brainer, let, let's not get into a sh a shouting match or a, you know, over which one is better. It, it's for what, what's the task?

And then it's very, should be relatively straightforward to say, this is better for that task, that's better for this task. Right? And we also, you know, there, the classical computers are all, we're, it's gonna be a counterpart, like even this, this experiment that we do, it's a lot of classical computation for post-processing and things like that.

Sure. To, to, you know, mitigate some of the errors that still exist in the system. So classical plays like a strong role in quantum computation, and we see that as going to continue, right?

We think there's always gonna be a handoff between classical and quantum, just like you said, even not only like what big picture problems, like what parts of the problem Yeah. Might be suited for quantum. And there's a lot of, you know, there's some, you know, simulating nature, right?

Quantum, you know, nature is quantum, quantum can simulate those aspects. Uh, there's a huge, you know, that's, that's lots of materials and drug discovery and all those kinds of things. Um, but there's lots of open questions, like optimization is something that we've seen a lot of movement on, but there are open questions.

Can it, can quantum do optimization, uh, in a way, you know, how much more efficient can it be an optimization, for instance? Uh, same thing with machine learning. There's a lot of evidence that data with like specific structures, uh, if you exploit a certain structure of, of these problems and of, of this data, quantum can lend a hand potentially.

But, you know, it's also the, the scale of the quantum device and, and how we continue to evolve. But like you said, I think the one thing to take away is that this is a tool that is going to be massively impactful for very specific problems. We need to continue to understand what those problems are.

And, and as you know, like discovery isn't planned, right? Th these early, the early things that we're doing, you know, we're looking at condensed matter problems, maybe like exotic materials, things like that, that will, we will learn more and we will find, you know, like it will, it will continue. We will find discovery along the way.

You know, I think sometimes we, like you said, we know that quantum's gonna be good at some things. We know that we kind of have places that we're looking. But to me, what's much more fascinating is like this blank slate.

You know, we, we have this kind of blank slate, slate of exploration where we're saying like, let's look at this tool. This tool does something differently than the other tools that we've had available. What can we use this tool tool for?

So Go where no computer has gone before. Yeah. That's an exciting part to me.

Yep. So, you know, I, I had a fascinating conversation with a fellow from NIST maybe two months ago, post quantum algorithms to protect digital certificates. Thanks.

You know, this kind of thing. So clearly, right? When you, when you look at the, the, the, at least in theory, the ability for a quantum computer to, to break 1 28 or even 256 bit encryption rather easily, you know, current encryption models.

Those, and, and granted that, that's not the best use of quantum computing in my mind, ive gotta be honest with you. But, but you know, clearly something like that, no brainer, right? We, we need that.

Great. And it also forces us to think about a post quantum world of like, we're still gonna need encryption and we, we wanna make it, you know, worth, uh, secure that even a quantum computer can't break. So, absolutely.

I I think there is that, you know, there's that, uh, cycle of, of ending in synthesis, right? Thesis and antithesis synthesis where we, you know, it, it's got the dog wagging the tail, but you have breakthroughs, you have, you know, then applications and new breakthroughs, and there's gonna be this, this whole laying, laying out. I think the world wants to know, though, Katie, when does it get real for us?

In, in not everyday life, perhaps, but you know, when you can really start pointing to some of these things that quantum computing are doing, or you can do it. Yeah. I mean, quantum computing is doing a lot today.

It's just this idea of like, when is it going to do something that is better than what you can do today or more efficient than what you can do today, right? Power, you know, resistant than we can today. Uh, I think if anyone tells you they have an exact answer on that, they're, they're not telling you the truth.

Uh, you know, it's all about scale, quality and speed. And finding these problems with these inherent, like the, the one that you brought up on the, the crypto, like factoring numbers. Ha it has these inherent, these inherent structures to the problem that quantum computers can exploit.

Um, so no one has the right answer on the date. I will say this, you know, the way that I say it and the way that I look at it is we put the first quantum computer on the cloud in 2016, it had five qubits, uh, you know, now, and it's not all about the qubit, right? You gotta have good qubits.

Uh, and our qubits, you know, all the qubits are getting better and better each day, but we were at five qubits then, like we're talking about the path to a hundred thousand qubits now with our partnership we just announced with University of Tokyo and University of Chicago. So the rapid advancement is, is mind boggling. Um, and now, you know, we're talking about a hundred thousand in the next, you know, 10 years, but we're already at 433, we have a chip that's gonna do a thousand.

We're talking past that. We're, you know, we're constantly gaining from the five qubits that we were at really recently. So the scale and the quality of speed is gonna continue to, to expand something I also think is really interesting.

Um, you know, when we, when we started understanding classical computation, we didn't have all the other tools that we have today. We didn't have armies of computer scientists and cloud computing and open source software. You know, the parallelization of, of growth with quantum is going to be very, very different than what we've seen before.

So, like, the acceleration is just, it's, it's gonna be really, you know, continue to be like, you know, um, spurred by all these other technologies running in parallel. And I, you know, I think even with GPUs, right? There was a long time, I think where we, we kind of knew what we wanted to do with them, what their hardware wasn't there.

And now we're looking, you know, how we're exploiting the GPUs, we're finding new things. You know, GPUs have been like the name of the day, right? Mm-hmm.

Nobody has a date. Uh, I would keep an eye on, you know, if I, if I'm looking at, you know, my business when to get invested. Um, another thing that, you know, I work with industry partners all the time.

Um, my, my team works with industry partners, everybody in your industry, I like, I'm willing to bet that they have someone who five or six years ago, quantum wasn't an industry, right? It wasn't commercial party. So if you were coming out of school with a, some background in Quantum and you didn't wanna go to an academic research lab or something like that, you went into industry, right?

You did something else with your skills. Um, everyone has someone in their organization that knows and enough about this technology to understand really where we are today and how fast it's progressing. So I, I, you know, my recommendation as we think about this, like, when can we time it out, is find that person internally that you know, you really can trust to help you follow the narrative, follow the advances with I b M, the rest of the industry.

Um, you have someone inside that I promise knows everything about this, you know, enough, least enough to be dangerous and more importantly, knows about your internal problems, right? They know your use cases. They know what your, where your limitations of your existing, you know, com, computational overhead is, and they can start to kind of map where we really are and where you need to go, uh, in your industry.

So the date is T B D, but it's fast approaching. It's coming. Well, I, I, I get the feeling also, look, you know, early on in these new technologies and, and new fields like this, it, it's kinda like the big bang of the universe, right?

You could go through that great expansion faster than the speed of light even, right? Where it just blows up. But then you reach a point where, you know, not diminishing returns, but you just, you just don't expand that quick.

I think Quantum is still in those first milliseconds, if you will. Uh, and, and you're gonna have this great expansion because you have all of these resources that you didn't have, you know, for computing 50, 60, 70 years ago. You, it's internet time, right?

It's the whole crunch of how, of how we do things here, and there's just so much to discover yet, right? Like, think about it, going from five qubits to, to a hundred thousand is, I mean, mind boggling when you, when you think about that, right? Um, it's not, I mean, you talk exponentially.

Yeah. It's, that's pretty exponential, right? So it, it's great stuff.

Um, Katie, for people though, who say, you know what, I'm fascinated by this. I wanna stay on top of it. Where within IBM can they kind of stay on in the know, on, on IBM's quantum, uh, breakthroughs in technology?

Yeah, we have tons of resources. I mean, from anyone from like a developer who wants to understand how to develop in quantum to, to industry partners who wanna really understand like where their business are and what the skills are internally. So we have like a, we have different offerings.

Uh, we have access offerings. We have a quantum accelerator offering that, you know, can help, uh, understand your team and the skills that you have on your team, and start to look at what a business case for your organization would look like in Quantum, what use cases might be valuable to you and how to get, but anybody can get started like today with the same browser. They're looking, uh, you know, they're, they're using to watch this video.

There's textbook, textbook educational resources. There's, um, lots of examples of how to run, you know, ground state simulations. Uh, there's, there's more information than you would want.

Uh, both, you know, both from an offering from IBM as well as open source, uh, community information. Love it. Katie, I appreciate you coming on and, and sharing with our audience today, this great, uh, story and information on Quantum.

You know, I always say I, I love what I do because I think our audience is like me and it's like talking to my people. And I, I'm fascinated and ever curious about, you know, quantum and, and all that it offers. So I'm sure our audiences as well keep up the great work over there at ibm.

This is kind of just foundational, kinda Absolutely. Research that's going on and, and needed. Yeah.

So thank you. We want this to be a scientific tool for discovery, and there's gonna be a lot to discover. Absolutely.

Excellent stuff. All right, we're gonna take a break here on text, on tv. That was Kate Pito, director, IBM Quantum Theory and Computational Science, but a new breakthrough from IBM and their partners at Berkeley, Kate.

We'll, Katie, we'll see you soon again, I hope. And have a great day. All right, bye-bye.

Bye.