Getting The Measure Of Quantum Computing - We Interview QUBT CEO Robert Liscouski

Summary

• Quantum systems are likely to provide a major leap forward in computing architecture.
• In its early stages today, quantum promises to work well in solving multivariate problems that are elusive, expensive or inefficient when addressed by classical digital computers.
• Such problems crop up in many computationally intensive end-markets, including financial services, pharmaceuticals, logistics and others.
• In our interview with Quantum Computing, Inc CEO Robert Liscouski we check in on the latest progress made by this exciting software company.

DISCLAIMER: This note is intended for US recipients only and in particular is not directed at, nor intended to be relied upon by any UK recipients. Any information or analysis in this note is not an offer to sell or the solicitation of an offer to buy any securities. Nothing in this note is intended to be investment advice and nor should it be relied upon to make investment decisions. Cestrian Capital Research, Inc., its employees, agents or affiliates, including the author of this note, or related persons, may have a position in any stocks, security or financial instrument referenced in this note. Any opinions, analyses, or probabilities expressed in this note are those of the author as of the note's date of publication and are subject to change without notice. Companies referenced in this note or their employees or affiliates may be customers of Cestrian Capital Research, Inc. Cestrian Capital Research, Inc. values both its independence and transparency and does not believe that this presents a material potential conflict of interest or impacts the content of its research or publications.

Introduction

In our CEO interview series we’ve been fortunate enough to discuss key topics with the leaders of Aerojet Rocketdyne (AJRD), Iridium Communications (IRDM), Science Applications International Corp (SAIC) and

Virgin Galactic (SPCE). We’ve covered turnaround management, deleverage, M&A as a growth lever and the challenge of operationalizing a dream as topics respectively. In this, our latest CEO interview, we turn to a new computing paradigm that we believe will be making its way into a datacenter near you sooner than you might think.

Quantum Computing, Inc (QUBT), the company, is a small cap software business currently trading on over the counter markets. It has a market capitalization of c.$247m (source – Seeking Alpha). Revenue is negligible, earnings negative, and the company is dependent upon equity raises to fund operations; it recently completed a c.$12m fundraise. None of which we would expect to set your heart racing.

But read on.

What Quantum Computing Is And Is Not (All At The Same Time)

Quantum computing, the paradigm, is a potential salve to the constraints that Moore’s Law and other physical factors – heat generation, energy efficiency - are beginning to impose upon classical computing.

For the uninitiated, classical computing is the world you know and indeed follows the rules of the world you can see, feel and interact with. It utilizes small form factor, highly standardized integrated circuits, typically manufactured using a silicon substrate (in more esoteric applications you might find gallium arsenide). These ICs contain many millions of transistor gates which can store a zero or a one value; binary code is the key to classical computing. When you click, “Like”, your computer translates that through a number of different abstraction levels until the CPU hears a binary string, and then outputs a binary string based on your instructions.

Advances in classical compute power have been a function of many things, but the number of transistors packed onto any one integrated circuit die has been one of them. The Moore’s Law of which many of us have heard, but few of us truly understand, was coined by Gordon Moore, founder-CEO of Fairchild Semiconductor and of Intel (INTC), and has to do with the number of transistors per chip – you can read more here. But you’ll hear Moore’s Law cited as a catchall for the mounting challenges in conventional computing, be it heat output, power consumption, ability to deal with complex uncertain multivariate calculations, and so on and so forth.

Quantum computing is not based upon ones and zeroes, but instead on probabilities and multiple simultaneous truths. You might think of classical computing as the Bauhaus of computing – the epitome of the Enlightenment Age. Linear to the last. And you might consider quantum to be truly postmodern in nature; facts and alternative facts being absolutely compatible within quantum logic. A quantum computer – yes that is a thing – doesn’t operate using transistor gates occupying only binary states – bits with value one or zero. It operates using ‘qubits’ with a value of one or zero or both. Qubits exist as energy states within superconducting loops in a quantum computer, the same way that bits exist as voltage at a transistor gate. Qubits can be a single value of zero or one – once a computing problem is solved, a qubit has reached its lowest energy state (which could be zero or one); but in the main occupy multiple simultaneous values.

If classical computing is ideally suited to linear tasks such as arithmetic, measurement, storage and so forth; quantum is ideally suite to nonlinear, multivariate and/or iterative tasks such as optimization problems. The simple question of route optimization isn’t really a simple question. The real question is multivariate. It is, “in what order should I deliver these Amazon parcels, given the weather, the cost of gas last week when I filled the tank, the degree of irritation each specific customer experiences when I am late, the proclivity of customers with irritation level X to switch from Amazon to other vendors, and the specific-customer profitability of each customer on my route”. The data – the variables – required to answer the question can be stored in any typical, classical database. But solving that problem efficiently and quickly? That’s better suited to a quantum device.

Right now quantum computing can be purchased; if you want to buy time on a quantum computer, a company called D-Wave will sell it to you in the same way IBM used to sell you telnet time on a mainframe. In addition – and this came as a surprise to us – you can also buy Qubits As A Service from, yes, Amazon Web Services. ‘Braket’ is the experimental service they offer.

Now, cast your mind back to the early days of commercialized linear computing. In the 1960s, you could really only buy software bundled with the hardware and sold by the hardware vendor. So if you wanted an accounting system, you might pay IBM for one, but it came with a pile of IBM metal that you probably didn’t want or need. You probably just wanted to get the accounts done on time with less errors. This vertically-integrated model slowed down computer penetration for many years; it wasn’t until the establishment of the merchant software model that computing truly began to proliferate. Software’s first meal was the hardware industry. As software ate into hardware margins, so the cost of computing began to fall – you can thank Moore’s Law for that too – and computer proliferation began.

Certainty Begets Uncertainty

In quantum, we expect the same to happen. Separation of church and state, hardware and software, is we believe an essential step to establishing this new computing paradigm as part of the future. And this is behind our interest in Quantum Computing, Inc (QUBT), a software business focused entirely on the quantum domain.

We talked to QUBT CEO Robert Liscouski to learn more.  The questions are ours, the answers, Liscouski's. 

Q. Can you introduce us to the team at QUBT?

A. Sure. My background is as a security professional, initially in law enforcement, then in federal Homeland Security and then CEO of a company called Convergent Risk Group. In 2018 I became CEO of Quantum Computing, Inc when I was asked to co-found the company by Justin Schreiber and some of his colleagues who had a keen interest in bleeding edge technologies. I have no background in physics but I do have a longstanding interest in technology – Convergent was and is a technology-based security service provider for instance, and I grew up programming in COBOL and FORTRAN back when punch cards were used for storage!

QUBT’s team really came together in May 2019 when we were able to recruit Michael Booth as CTO and Steve Reinhardt as VP Product Development. Michael joined us from D-Wave Systems, a quantum computer systems builder, and was previously at Cray Research. Steve was also at D-Wave leading the software applications side of that business, and previously was at Silicon Graphics and Cray Research. In Michael and Steve we have two leading experts in the field, with sound commercial background, driving our core technology and products.

Q. What are the cost- and performance- promises offered by quantum computing?

It depends on the application. For regular computing tasks, classical computers remain cheaper and faster tools than quantum. But for the tasks that governments and enterprise users are starting to demand computers deal with – eg. intelligence agencies anticipating which individuals may go rogue first, weather forecasting agencies trying to work out the most likely path of a hurricane – we believe quantum offers fundamental advantages over classical. Whenever the answer to a problem isn’t binary but probabilistic, that’s where the possibility of quantum beating out classical lies. If you think about it for a moment, there are a huge number of real-world problems that are probabilistic in nature. Logistics optimization, securities trading, drug discovery – all these domains require answers to problems that either have no absolute answer (which is the absolute best parcel route?) or which have answers that are impacted by the first step in solving the problem (how do I invest $1bn in a particular stock on behalf of my clients without moving the stock so much that my last client has to pay a lot more than my first client?). In these worlds, quantum computing can we believe provide better answers than classical.

Today it costs more to process a query in a quantum environment than a classical environment. That may change as quantum grows in scale and popularity. If we achieve anything like the cost improvements obtained in classical computing, we can imagine that some problems will not only be better solved in the quantum domain but also more cheaply. Today the cost is high because the hardware used is bespoke. You can’t buy a quantum computer from Dell, and they don’t look like a blade server or anything else you will find in a classical datacenter. Early machines looked like fusion reactors, and current models look like large standalone supercomputers. And that, in part, is because quantum computing can only operate at temperatures close to absolute zero. Quantum machines today are still very early in their evolution.

Q. Will enterprises wishing to use quantum have to undergo a wholesale migration, like mainframe to mini to client-server to cloud? Or can quantum be slotted into the datacenter somehow?

The latter. Quantum computing isn’t a platform designed to replace classical computing. There are tasks that classical is better suited for and likely always will be. But there are tasks that naturally lean towards a quantum approach. We see the near-term future for quantum computing as making available quantum compute cycles to customers for the applications that they want to run in a quantum environment. Today for instance, you can buy experimental quantum services from Amazon Web Services – the ‘Braket’ offering. In the future we anticipate that, for instance, workloads sent to AWS will be handled by classical or quantum computing devices depending on requirements.

Q. Where does QUBT sit in the value chain and why?

Our goal at QUBT is to see our software proliferate in enterprise computing production workflows . Our control plane software sits above classical and quantum computing devices, receives incoming workloads by way of a single API, and can then define the problem in quantum or classical terms, passing the problem to the appropriate device. Once the problem is answered – let’s say it’s a route optimization and we now have a successful route generated – we pass it back to the application which sent the request. The idea is that the API call from the third party application can be written by any skilled developer. You don’t have to be quantum-schooled. This means that instead of each developer spending around eight months getting familiar with, say, IBM’s quantum developer toolkit, they can express their problem in a quantum-ready form via our software in a matter of days. In essence we provide software which accelerates and makes it cost-effective to run applications on quantum computing platforms.

Q. What stage of commercial development are you at right now?

QCI is just getting into the market place with its solutions. We have spent the past 18 months developing, refining and testing our software applications and platform with various beta users to ensure that we had a practical and workable software platform before going to market. As we discussed earlier, there are a huge number of real-world problems that our software platform can address; logistics optimization, cyber security, drug discovery, etc. Our initial focus is on the complex problems that we are certain we can solve and scale into a broad market, such as supply chain and logistics. We have just completed our initial hire of our sales team and our goal in Q1 and Q2 this year is to acquire clients that are willing to partner with us to demonstrate the real value of quantum computing in their business environment.

In fact, the development of real world use cases is so critical that we have initiated an internally funded program called QikStart which will provide access to QCI’s industry-leading quantum acceleration platform, expert resources, and funding to explore and push the boundaries of quantum computing for delivering practical business results, right now. QCI’s Chief Technology Officer, Mike Booth, will be leading the QikStart Program. Mike’s deep experience and understanding of how quantum computing can add value to critical business solutions is a significant advantage to companies interested in participating in the program.

Perhaps most importantly, we are also connected into AWS Braket and can connect into any of the three quantum computers, those made by Rigetti, DWave and IonQ, right now to run applications. QCI is the first company with a quantum acceleration software application and execution platform to provide the widest selection of QCs via the cloud, which we believe will be the most practical and scalable approach to bring quantum computers to business users.

Q. What does the quantum vendor ecosystem look like today and how do you think it will evolve?

The quantum computing ecosystem today is in a very early, embryonic stage. There is significant investment in the development of quantum hardware by governments and the private sector world-wide. This is because it is widely believed that the nation or company that achieves real quantum advantage will bring significant competitive advantage to the industry. But, as you have pointed out, true quantum advantage is years away. On top of that, this is a technology that requires significant investment to deploy; initial capital costs to acquire a quantum computer, specialized infrastructure to maintain cold temperatures, significant operations and maintenance costs, and specialized programmers to keep the machines running, tuned, and ready to run applications. While national governments will purchase quantum computers (similar to classical supercomputers when they debuted), it is envisioned that few companies will have the business need to justify the budget required to support a quantum computer of their own. Most software vendors that we consider competitors are focused on lower level programming and a bevy of consultants to address programming and application development for business and government users today.

However, we believe that quantum computing in the cloud, similar to the model that Braket is offering is the trend that will continue to grow. This is where QCI is focused with our platform approach. Developers can now exploit and evaluate different quantum processing units (QPUs) simply by selecting the target QC from within the our platform interface, which we are currently calling Mukai. Mukai’s software execution layer can also direct the execution of problems to either classical (Intel or AMD processor-based) or a hybrid of quantum and classical computers, based on user guidance. Mukai users do not need to learn or implement the varying highly technical methods of connecting the classical environment to the target QC. By providing a common, simplified interface, Mukai can save developers and programmers significant time and resources, while vastly accelerating the development process. And most importantly, users can run their applications on any of the three quantum computers available in Braket; Rigetti, IonQ and Dwave. The advantage of doing that goes far beyond the time and money savings to gain a performance benefit, but the users will be able to see which of the quantum computers performs best for their problem – futher optimizing the outcomes. This is a very exciting approach for QCI and the user community.

Linear Progress

You can learn more about Quantum Computing, Inc at Quantum Computing Inc.,

We will cover QUBT later in the year once it completes its move to the Nasdaq main market.

Cestrian Capital Research, Inc – 22 January 2021