Episode 6: Quantum Computing Supremacy (Part 2)

News & More /

Featured


Episode 6: Quantum Computing Supremacy (Part 2)

In part two of our most recent podcast episodes about quantum computing, on The Disruptive Enterprise Podcast, host Greg Turner and guest Bill Kleyman discuss just how far computing has really come. Today, an entire server farm can perform AI, machine learning, and neural networks all powered by GPUs. Who saw that coming? In part two of our podcast, Greg and Bill discuss the recent news about Google, IBM and quantum supremacy and what that might mean for the enterprise.

Has Quantum Supremacy Finally Been Achieved? (Part 2)

With Gregory J. Turner & Bill Kleyman

Gregory J. Turner: Welcome back to our continuing podcast series at The Disruptive Enterprise. Today we are talking about quantum computing considering some very new and recent information. With me to help discuss this topic is Bill Kleyman. Bill, as a brief reminder, is an advisory board member for MTM Technologies, and he is the Executive Vice President of Digital Solutions at Switch.

Now here is Part 2 of our discussion.

Bill Kleyman: I want to take a step back. And I do have a couple of other thoughts around this quantum computing conversation that we’re having. We don’t have to answer this question here, but what is, in today’s day and age, approaching 2020, a classic computer? You’re talking about Cisco, their UCS designs, now, more integrated with these cool and video grid cars that can do more than anything before.

Bill Kleyman: And on that note, a classical computer has been, let’s be honest, redefined, by what this GPUs are doing getting entire trace and arrays of full GPU processors, an entire server farm that could do things like AI, and machine learning, and neural networks all powered by GPUs. Is that a classical computer these days? Is that some other things? It’s completely left field. Who, 10 years ago, would have believed you, maybe even 15 years ago, if we said, “I’m going to power an entire AI engine with graphics processors”? They would have laughed at you. They would be like, “No way. Intel is way better than you,” right?

Gregory J. Turner: Right.

Bill Kleyman: But that’s the case. That’s completely redefined the way we do some of the most advanced computational processes in the market. This is just what happens with GPUs. Look at what HP has been doing. Moonshot, for example, with its high performance, highly parallel processing architecture that’s capable of doing extraordinary things or things at computational science, parallel processing, and doing things like advanced research.

Bill Kleyman: So, I think that everybody, sort of, listening and paying attention to this industry needs to take a note that classical computers have now changed from your traditional box with a CPU in it, with some RAM, and a hard drive. They’re much, much different. And I think they’re much more use case-specific. So, when compared to classic computers, quantum machines—I’m going to quote a part of this article that I read from the American Association for Advancement of Science. In that article, they talk about the use case that you and I are talking about, where quantum computers are not necessarily supreme against classic computers because of a, “laboratory experiment designed to essentially implement one very specific quantum sampling procedure with no practical applications.”

Bill Kleyman: And maybe I agree to some extent. And the other fact of the matter – and we’ve talked about this, Greg, and everybody listening – is that these quantum machines need an ability to do error correction. And the Google computer lacks the ability to correct errors, which, in my opinion, maybe potentially the last key to making it a full-fledged quantum computer, where you can actually require a stable, logical compute state that can do things like error correction. So, in my eyes, Greg, we saw an example, a potential, a capability here of a machine that operated for about 200 seconds and solved a problem or maybe a little bit of a challenge that didn’t really have any sort of practical real-world experience, and it really wasn’t a machine that did something cool, but it’s got a long way to go.

Gregory J. Turner: Right, right. Well, and I think your point about what’s really considered a classical computer today as opposed to what was one—when I was starting out in the business world, it was literally an 1888 chip, what is our personal computer. And I had the IBM PC, and XT, and then AT. Holy cow, 10-megabyte hard drive. So, from those days to where we are today, it’s hard to compare the classical computer today. And really, it’s got a lot of advantages. And certainly, the way it processes with the RAM, and all the flash technology, and drives, and solid-state components that are available today that weren’t available back then.

Gregory J. Turner: So, I’m okay with having something in a laboratory that’s only available in a laboratory. And I’m okay with something that still has error conditions or error checking issues because from that is a building block. And from that building block, we can now start to expand the research and development, and we can start to create those error checking qubits, if you will, to improve the results and accuracy, and we can improve the environmental conditions by which these things operate.

Gregory J. Turner: I think for me, the explosion of opportunity was the fact that we weren’t really dealing with algorithms anymore. We were literally dealing with an input process, output result, which is really the basis of computing. And from that, Google generated a result, which was dramatic. And so, I think it’s all positive. And from the guy that likes to read about research, I think, for me, it’s certainly great news, and I think it deserves the flag-waving.

Bill Kleyman: So, let’s do that. Let’s wave some flags. I’m certainly excited in progress. And there’s going to be a lot of interesting things that are going to come out of this. It’s good reason that Google’s achievement was documented by more than 77 authors in a prestigious peer-review journal. It’s not surprising that this is something that’s being worked on. I think quantum computing ideas that Google’s been actively applying for more than 13 years, this isn’t something they just picked up in their backyard.

Gregory J. Turner: Right.

Bill Kleyman: And they’ve got a lot of really good use cases in mind. Doing things like complicated optimization problems, like calculating how to deliver packages in the shortest amount of time with least energy, improving encryption technology, building machine learning systems, or like doing things like simulating real physics of molecular scale materials. I mean, there’s going to be some cool things that they’re going to be doing here soon.

Bill Kleyman: There’s certainly limits currently, at least, to what they’re doing right now, some of the problems that they’ve solved through this error correction thing that they’re working on. But let’s talk about Google’s quantum optimism here. They’re not even trying to go after Moore’s Law anymore of classical computer exponential growth. They’re trying to more than double it. But they’ve got a long to do list. Improving how long these qubits can run error-free, maintaining that quantum coherence state, and maintaining quantum entanglement. That’s their number one right now. If it was my opinion, that’s what they must do is improving the errors of the actual device.

Bill Kleyman: And then, later, that’s when you’re going to see some true fundamental changes when quantum error correction techniques can be used to sidestep some of these qubit instabilities. But it’s going to happen. And I think they’re on the right track. I think Google’s being quantumly patient, and they’re certainly trying to go after that market. I mean, listen, I’m excited. I think that they deserve to wave this flag. It’s like mind-bending behavior, and the atomic scale of physics trying to create when it comes to quantum computing itself. I mean, it’s really, cool, everybody. And just don’t think this is going to happen tomorrow. There are still years of research and, surely, development. What’s interesting, Greg, and I want to make sure everyone’s aware of this, is that for those who want to try out Google’s quantum computer, the company is planning to make it available via their cloud service in 2020.

Gregory J. Turner: That is cool.

Bill Kleyman: Pretty cool.

Gregory J. Turner: And I, definitely, will want to be one of the guys trying that out. Before we kind of wrap up on this topic today, one of the things that strikes me is the conditions of errors. And I’m always reminded about the story of if you didn’t know what the probability was for flipping a coin, you might suggest that its heads would only show up 30% of the time, or 40% of the time, or 25% of the time. But because you flipped the coin several times, say a thousand times, you finally realize that it’s 50% of the time. And then, kind of the notion of heuristics.

Gregory J. Turner: And so, I think with quantum computing, and I’m certainly no scientists like you, Bill, but I think with quantum computing, clearly, there’s an opportunity to include the heuristics of the error conditions, so that it’s not so bad to have an error because the error itself can point you in the right direction. And I think that gives us a lot of promise for quantum computing as well.

Bill Kleyman: You know, you’re not wrong. There is a need to improve quantum systems that still lack error correction, meaning that they won’t necessarily be completely sure of every calculation that they make. But you’re right, researchers are still saying that amid these mistakes, you can still have some extreme power, like whether you’re calculating something that happens in nature, these systems are going to potentially make mistakes. And sometimes, you can still, like you said, gain a deep understanding of where to approach and how to approach a certain calculation, or a certain variable, or a challenge, even though there are some mistakes in there.

Bill Kleyman: So, I agree. I just think that we need to take some of that variability and some of that challenge as far as the unsureness of how these calculations are being made, how accurate they are. Out of the equation, of course, if a machine is completely beset by errors, that’s not going to be very useful either.

Gregory J. Turner: No.

Bill Kleyman: It’s exciting times.

Gregory J. Turner: Yeah.

Bill Kleyman: It’s exciting times.

Gregory J. Turner: If it was trying to calculate my financial statements, then we probably want to make sure it’s operating correctly every time.

Bill Kleyman: That’s a good point. So, note to self, don’t use these quantum machines for Greg’s financial statements. Got it.

Gregory J. Turner: Bill, any other thoughts you’d like to share with us because this has been phenomenal?

Bill Kleyman: There was an interesting point back in—when was it? I think it was 1981 by famed physicist Richard Feynman, who described, in 1981, how these machines are moving into reality. And he said, “Nature isn’t classical, damn it. And if you want to make a simulation of nature, you’d better make it a quantum machine and quantum mechanical.” It’s a speech he said over at MIT. And you could approximate nature with a simulation in a classical computer, but Feynman, what he said, wanted to create a quantum computer that offers the real thing, a computer that will do the same thing as nature.

Bill Kleyman: And we’re going to get to that point. Whether it’s IBM or Google quantum computing machines, we’re going to get to a point where this hunger for better simulations, even sometimes with errors, are going to become a true reality, where we can not only mimic things like nature but also like drug compositions, genomic sequences, financial statements, maybe not yours, Greg, but maybe in general to make better decisions and predictions certainly. That’s a beautiful dream. And even from this 1981 speech at MIT by Feynman, there’s beautiful aspirations and goals to what I think Quantum Computer can do.

Bill Kleyman: And when we get there, when is the wrong word, I just realized. I don’t want to say—actually, I was going say, if we get there. When we absolutely get there in our progress as people, because people are super curious and would like to press those buttons, it’s going to be an amazing time. And I certainly can’t wait to see the kinds of things, the kinds of lives that we can save, and the ways that we can improve people’s lives by using quantum computing.

Gregory J. Turner: Bill, thank you so much. And thanks for helping to break this new and important finding from Google and IBM. And to all listeners, thank you for being with us today. I hope you found this podcast helpful. For any questions or comments, please feel free to send me an email at gturner@mtm.com. And for more about MTM, please visit our website, mtm.com. At the Disruptive Enterprise, this is Greg Turner. Thank you.

News & More /

News


News & More /

Perspectives


News & More /

The Disruptive Enterprise Podcast

Leave a Reply

Your email address will not be published. Required fields are marked *

Name *

 
Find Rewarding MTM Careers
See Our Latest News and Events
Get to Know MTM
Find Rewarding MTM Careers
See Our Latest News and Events
Get to Know MTM