Quantum Optimization, Hardware is the Biggest Obstacle

Interviewee: Davide Venturelli
Associate Director for Quantum Computing of the Research Institute of Advanced Computer Science at the Universities Space Research Association (USRA)
Interviewer: Radhi Shah
Foreign Attorney, Zuber Lawler

Details

Transcript

WE NOTE THAT THE FOLLOWING TRANSCRIPT WAS CREATED BY A ROBOT SO PLEASE FORGIVE ANY TYPOS.

[Radhi] Hello and welcome to our dead cat live cat interview series. DCLC is an online content computing magazine founded by a bunch of nerdy lawyers. But we are also scientists, engineers and futurists focused on the likelihood that quantum computing coupled with AI are going to fundamentally change our world within two decades. I am Ravi Shah, a foreign attorney with Zuber Lawler specializing in technology law and protecting IP rights. My work entails working closely with technology, computer technologies and the healthcare sector. Zuber Lawler is a law firm with offices across the United States in seven different cities doing major work in traditional areas of technology. And we are exceedingly excited about emerging areas of technology that will drive the economy 10 to 20 years from now. Thank you for joining us. I am very excited for this interview as we have Dr. Davide Venturelli with us today. Davide, welcome. Why don’t we start with introducing yourself and let the readers know a little bit about yourself?

[Davide] Sure. Hello, everyone. Glad to be here. I’m Davidow interlay. I’m Associate Director for quantum computing at the university Space Research Association, which is an entity that exists in 1969, when NASA brought back the rocks from the moon, and then all the universities wanted to study them, and NASA never thought to work with universities. So they created an association to start creating programs that bridge the gap between academia and the space agency. And then this was like, you know, more than 50 years ago. But now basically, we do a lot of collaborative research programs with in this field of aerospace, and artificial intelligence and the likes, not only for NASA, but also for the Air Force for DARPA for the NSF, or for the US government in general. But, you know, as you know, the motto of NASA is for the benefit of mankind of humanity. So for everyone, I suppose. And I lead, you know, the quantum computing program, which has a lot of senior researchers involved, that, you know, work for NASA and for other projects in quantum computing, specifically.

[Radhi] That’s pretty interesting. So how did your journey start in quantum computing?

[Davide] So you know I I used to like physics as a subject and at some point we started reading a lot about both string theory, as well as. analytics primarily because of the curiosity of trying to understand how how reality works and even similar to any any physicist, although I was not one. And then in 2017 when I was working with Microsoft, Microsoft announced that they’re working on a to sharp. Programming language for quantum and that gives me an opportunity to go deeper into the subject to quantum computing and I got more deeply involved, and since 2017 have been extremely involved in the point of ecosystem. yeah so that’s that’s how it happened.

[Radhi] I know you said you founded your company, but insight and advisors what does he do.

[Davide] Well, I, I think that my earliest appearance on the web, I was a teenager and I, I wrote, I wrote in a bulletin online somewhere. What are the universities where to study quantum computing? And it was, I don’t know, 1994, or something like that. And so I always been very fascinated about No, it was 97. Okay, who couldn’t be before that? Yeah, no, I was in high school though. And I really been fascinated by the idea of using of grappling physics and quantum information. And, and since I was a kid, I was undecided between either studying computer science or physics. And I thought theoretical physics in the end was the right idea. But I did most of my studies in this direction. And when this field of physics became an, an industry or started to become an industry, I was very fortunate because I was, I was already essentially a postdoctoral scientist at NASA Ames Research Center. And so I could I could witness from the epicenter of the cyclone, the unfolding of a lot of initiatives that that created the principal stakeholders of quantum computing today. So none of that’s an answer. But my journey started in Italy, and then I did a PhD, and I did publications initially in condensed matter physics. And then when I moved to the United States, I started working for the for government projects, essentially.

[Radhi] That’s amazing that you knew what you wanted to do since you were a kid and you follow that throughout. So very inspiring. Um, so like, just to familiarize everyone tuning in, can you give us your explanation of what is quantum computing? And what does your research and tell right now?

[Davide] Yes, so a little bit like a sad one to computing is is, is a phase on on the intellectual journey of, you know, humanity in some sense where we are really trying to see what we can do with the most fundamental laws of physics that we know how to formalize and how to use them to either you know, transmit, or process information essentially, and, and, and, and the idea is to, to build a programmable device that is able to exploit all these effects that happens in the nanoscale from, you know, tunneling superposition entanglement, and, and codify them in in some sort of language that is able to be expressive enough to write computer programs. And this idea has been around since a few decades, but only in the last decade has been embodied, implemented in real architectures that are sufficiently engineered to be, you know, to be viable at scale. And quantum computing is this attempt that we are doing to create something extremely sophisticated, keeping the quantum information alive and manipulating it precisely in order to implementing algorithms. And there is a theory aspects of because the great thing about quantum mechanics is that it’s not digitally simulated in an efficient way. So we cannot, like we could do for a standard computer, really write a computer program that simulates exactly everything that we want to design. So we need to do a little bit of trial and error. And at the same time, that is why we want a quantum computer in the first place, because we know that the quantum computer will be not replaceable by a non-quantum computer. And is that a sufficiently concise answer? I know that your audience has been listening to a lot of quantum computing experts to being the so I think that, you know, everybody knows what’s a qubit? Or what are the basic algorithms that are out there? But I offer the more philosophical answer.

[Radhi] And that’s, that’s correct. I mean, if, of course, it’s always good to know something from the person who’s right in working in that field and his theory on what quantum computing is, right? So yeah.

[Davide] You asked what my research entail, because I didn’t answer that part. So, my initial start as a researcher in in quantum computing specifically was oriented, still is in some large part, in to what is could be called Quantum optimization, which is a framework for which you are asking yourself the question, if you have the entire toolbox of quantum mechanics and you’re trying to identify the optimal solution of a complex problem that requires a search among an exponentially manual factorial in many candidate solution, how would you use quantum mechanics to do that? And to some extent, this is a problem that could be mapped to a cooling process, you have like some sort of like super complex material where every configuration of its internal degrees of freedom represents a candidate solution, but there is one of those solutions which represent the most stable state of this material. And to some extent, nature has very efficient ways to deliver to you very stable phases of matter. And we’re gonna have like a full understanding of how, you know, all the tricks of nature really achieve, for example, efficiency in some of these processes. But by creating the quantum optimization algorithm, we are setting ourselves to this challenge. And I were saying, Okay, now let’s, let’s take a problem in industry, whatever it could be, you know, logistics or in, in finance. And let’s map it into a complex interaction between spins, which are two dimensional variables. And then let’s find a way to drive a quantum evolution of this spin system into a state, which is the most the least energetic state of this couple interactions. And this is a very interesting physics problem. It’s a problem, which is, you know, time dependent disorder interacting. It has all the difficulties that you could study in the textbooks. And that’s the problem that quantum computers tried to solve either through gate model approach, through analog annealing approach or through other means. And this is a large part of my research. And I’m trying to invent new methods and test and evaluate methods of others on real device to try to achieve this kind of results.

[Radhi] Right? So what are some of the major technological obstacles that quantum mechanics must still overcome?

[Davide] Well, there are four. Well, they’re more than four, but there are like four leading technological paradigm that have been subject of billions of dollars of investment worldwide, on how to create a quantum computer, and probably there will be five, six, seven, or eight years from now, because there are emerging ideas that are becoming industrialized. But But today, you could say like, you could try to create a quantum computer with a with a superconductor, superconducting circuit, or with ions in a track, or with neutral atoms in optical tweezers, or via photons, which travels in some interferometers or interferometric devices. And this four paradigms, they have all challenges and pros and things, none of these store really has shown that is going to be the technology that will really dominate each other, the others. And I believe that, you know, advancing these hardware technologies is the major challenge that we face, because algorithm people like myself, at the end of the day, are hostage of the hardware that is built, we can devise wonderful ideas, but then at the end of the day, if we cannot run them in a real world device, the impact is not as large. And to be fair, especially, I mean, certainly in my field of research, where I’m not like a super mathematical geek. I care also about heuristics? ( Don’t know this word) , which are those algorithms that have no provable guarantee of success. And they need to be, you know, proven in a quick and dirty or messy way, in real world devices, and to some extent, like the entire field of deep learning and data driven methods. Has this kind of feature or flaw that you there’s very little things you can prove, without really trying. And so the challenge is, we need this Harvard technology to improve to have the required fidelity scale noise control. You know, there are some of these technologies they are a viable roadmap towards error correction, which would enable those devices to act as idealized quantum computers. Some other technologies they have not that roadmap as clear, but they still have benefits of being scalable and subject to error mitigation techniques, which would still make those devices interesting. We can go in details in each of them. But I believe that the major challenge we face is in hardware right now, we really want to see this devices to be a, you know, sufficiently large number of qubits and sufficiently precise in noiseless than some extent, so that we can use them to do things which are not simulatortable in any other way. And they are legitimately in a regime of what was called supremacy or superiority with respect to digital computers.

[Radhi] So when do you expect this technology and the applications of this technology to be profitable, and by how much?

[Davide] The most optimistic CEOs out there, they give, you know, 24 months’ timeframe for clear quantum advantage in business applications, most conservative ones, they, you know, tentatively, say, a decade. I, you know, I think that there’s reason for optimism, and that we will see signatures of an interesting advantage within a couple of years. But, you know, for this technology to really go in production and be embedded in industrial processes, this is not something in my view that we will see, you know, this decade, more likely than not. It requires system integration, or a lot of things that we will not, we will not be focused on at the moment. However, I might be wrong, I’ve been wrong many times, and, coming from the scientist world, as opposed to from the business world, I tend to be a little bit more conservative in my estimates, and I’ve been so impressed by the talent of entrepreneurs that have decided to be trailblazers, and even if, unfortunately, the performance of the actual quantum processors that I see today is not impressive on an absolute scale, like I wouldn’t use them to solve a business problem today, they are to some extent, most of them are really impressive from engineering scale, like, I would not have predicted few years ago that this level of sophistication and so its difficult for the human mind to interpolate exponential trends, but I can totally see things becoming very valuable, especially in the scientific R&D field in the next couple of years. Meaning that applications which are more you know, self-referential to some extent, like how to calibrate quantum mechanical systems, how to simulate non equilibrium, models of condensed matter, these kinds of things, which are more naturally fitting the actual working of a quantum computer are probably going to be impacted very, very soon. The science market is not necessarily the what we’re going after, but after that, once you keep those things in control, once you make them reproducible, once you make them well calibrated, and you can build on that and the next industry could be you know, finance or machine learning as a technology track. So I don’t know how profitable all of this will be. I believe that the market eventually will be very, very large because it’s, it’s an upgrade on essentially every aspect of computation, that it can be done. And if you add to the fact that to computing you have sensing you communication that it’s an upgrade to the entire information technology world. But, but to realize that large revolution or large scale impact, again, not this decade, in my view. But I think there is the case for some industries to try to be protagonist of this development. Because early mover advantage is a thing if you’re a large company, and I believe that, you know, if you’re in, in a business like oil and gas, or if finance were having this headstart might pay off big, it’s a good idea to start investing a bit in this direction. One thing that I would say to us also is that specifically, if you’re looking at software innovations in quantum computing, they often come with innovation in the digital computing for free in some sense. By trying to exploit physics effects and building a quantum algorithm. It has been done over and over again over the last few years, that you invent or innovate classical optimization methods, or classical machine learning methods, with techniques that were not known before, and that they are beyond state of art. So even if the end game is to run these things in quantum computers, and hopefully to gain some something out of that, the journey is also generating ideas which are viable today, on devices, which are known quantum, the entire buzzword of that, which is the quantum inspired methods. And there’s a lot of very skilled research groups that are focusing on that, and in technology platforms, which are admittedly non quantum, but they have some elements of working principles that have been developed because of the research in quantum computing. And I’m working on some of them, and I’m impressed by their performance.

[Radhi] Right, so, Dr.Venturelli, if you would invite three scientists past or present to our dinner party, who would you invite? And why?

[Davide] Well, if they’re present, I would just write them I suppose. I’ll focus on the on the past. I actually don’t know if you ask this question to everyone. I don’t know what the answer but I would be surprised if people would not invite Richard Fineman considering that is, you know, kind of the or the most fun and wise physicist in the in the history of the field and is also very good Bongo player. So it would be you know, good reason just to you know, spend an evening with a guy. I would invite people that have secrets so that they can tell mesecrets that were lost in time. Like Thermat? ( No Idea who this is ) that wrote, you know, the little note that people have been wondering whether you had a method to solve that problem, which didn’t require the hundreds of pages that has been done recently to solve the last theorem of thermat or for I would invite Leonardo da Vinci so you can it can take you know, to do a portrait of myself I can sell for hundreds of millions, I suppose.

[Radhi] Yeah. So, as we are nearing the end, is there anything you’d like to share with our audience today?

[Davide] I’m glad that you guys exist and disseminate good ideas and interesting characters that talk about their work. I think this is a very fortunate time for people like me and for geeks in general, because we have grown the attention on a thing, which is, you know, very, very cool. Quantum computing and quantum technologies in general, in my view, are a type of technology which has just a level of nobility, which is stronger than most of the others, okay? I’m biased of course, but you know, go blockchain or those are derivatives technologies like that or not, the intellectual merit of those technologies, which are very important for the future of our planet, like 5g or something is great but is derivative is somewhat like, heavily rooted in engineering and not into a fundamental quest for knowledge, you have to understand that quantum physics might not even be the fundamental theory of reality, that we have challenges that we were embarrassed of not understanding of how the universe works, why there are, you know, why we don’t see dark matter or dark energy, and maybe the mathematical framework we have is not the most correct and by creating an industry that is based on those, this mathematical framework by forcing yourself to create devices that have to function in a very intimate way with our understanding of nature, we will also advance our understanding of nature itself. We will discover things about the character of fundamental particles of, of laws of physics that, you know, we would not if we didn’t have these tools. So I’m very excited about this. Maybe nonbusiness application of quantum. And I believe that that’s reason enough for the 10s of billions of dollars that the governments are putting to subsidize this industry. I’m very grateful that big tech as well as young startups are focusing their energies in trying to advance this field as opposed to try to, you know, learn get a new way to optimize advertisement, the you know, website or something like that. So, the thing is very, very lucky. You know, kind of final message.

[Radhi] Thank you so much for your time and being on our platform here today. It was an honor for me to interview today, for sure.

[Davide] t was a pleasure for me to be here.

Davide Venturelli - Guest

Associate Director for Quantum Computing of the Research Institute of Advanced Computer Science at the Universities Space Research Association (USRA)

Davide Venturelli is currently Associate Director for Quantum Computing of the Research Institute of Advanced Computer Science at the Universities Space Research Association (USRA). He has worked since 2012 in the NASA Quantum AI Laboratory (QuAIL) under the NASA Academic Mission Service, invested in research projects dealing with quantum optimization applications and their implementation in a hardware-software co-design approach. He has authored more than 40 publications and 9 patents on the subject of AI, Theoretical Physics, Quantum Computing, and Robotics. He teaches Quantum Integer Programming as an adjunct professor at Carnegie Mellon University Tepper’s School of Business. He has experience winning, managing, and leading multi-million dollar R&D projects as Principal Investigator or co-PI, sponsored by DARPA, NSF, and DOE. He is the co-lead of the Ecosystem task of the National Quantum Initiative Superconducting Quantum Materials and System (SQMS) Center at Fermi National Laboratory. In 2021 he was elected member of the Quantum Economic Development Consortium (QED-C) steering committee, the organism coordinating 100+ companies involved in building the supply chain for the emergent quantum technology industry.

Radhi Shah - Host

Foreign Attorney, Zuber Lawler

Radhi Shah focuses on commercial litigation, particularly patent litigation, as well as patent prosecution. Radhi developed her legal acumen at various international law firms, often representing clients in computer technologies and health care sector. She has worked with a wide range of technologies, including cloud computing, machine intelligence, embedded systems, biotechnology, life science, and medical devices. As an Indian Patent Attorney, Ms. Shah has worked closely with wide-ranging international clients, forging connections between U.S. and foreign entities. She likes to get “under the hood” to gain a deep understanding of her clients’ inventions, allowing her to develop comprehensive patent strategies.

Leave a Reply

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

one + three =