How Quantum Computing Software Gets Built
Quantum computing is kind of like your middle-aged alcoholic uncle who’s still waiting for his career in the major leagues to take off. Every time you talk to him, he’s almost managed to get that big break which is just around the corner. Even though you really wish it would finally get here, you’re kind of tired of hearing about how it’s “just around the corner,” and deep down inside you suspect it’s never going to happen. That’s sort of how we feel about quantum computing today, yet the news continues to pour in. The first question we find ourselves asking is, what exactly are we waiting for?
Quantum Supremacy and Quantum Advantage
IEEE Spectrum published an article last month titled “When Will Quantum Computing Have Real Commercial Value” which talks about a committee assembled by the U.S. National Academies of Sciences, Engineering, and Medicine that was tasked with answering that question. You can download the 206-page report with the findings of the committee or you can just read the following paragraph from the IEEE Spectrum article which distills the report’s findings:
The committee concluded that “it is highly unexpected” that anyone will be able to build a quantum computer that could compromise public-key cryptosystems (a task that quantum computers are, in theory, especially suitable for tackling) in the coming decade. And while less-capable “noisy intermediate-scale quantum computers” will be built within that time frame, “there are at present no known algorithms/applications that could make effective use of this class of machine,” the committee says.
People often refer to “quantum supremacy” as the point when we are able to do something with a quantum computer that we can’t do with classical computers. There’s also another term being floated around lately. “Quantum advantage” is defined by IBM as “the point at which quantum applications deliver a significant, practical benefit beyond what classical computers alone are capable.” They’ve also come up with their own way to measure their progress, something they’re calling “quantum volume.” The company’s expectation is that they’ll reach quantum advantage within the decade. Ask any IBM shareholder how they feel about IBM’s ability to forecast and you won’t get many votes of confidence for charts like this one:
A lot of what many of these companies talk about seems to be around public relations. IBM’s most recent “AI breakthrough in quantum computing” was all over the media, but when the press release ends in “we are still far off from achieving Quantum Advantage for machine learning” it seems to be more about keeping up public appearance.
The question remains. With all the efforts being made on the hardware front, how far are we away from being able to do something – anything – useful with quantum computers? To answer that question, we traveled up to Vancouver, British Columbia and sat down to talk with Landon Downs, Co-founder and President of one of the world’s leading quantum computing software companies – 1QBit.
With just over 100 employees, 1QBit is a fast-growing company that focuses on using special purpose computer hardware from third parties to improve upon solutions to computing problems where it’s increasingly difficult to squeeze out gains in performance. Their platform is hardware agnostic and is designed to be able to use any sort of new computer hardware whether that’s quantum computing hardware or not. They’re not just sitting around waiting for whatever it is we’re all waiting for. Instead, they’re already working with some big clients to solve some big problems.
A few years back, 1QBit began working with Dow Chemical to develop quantum computing applications for materials science and to improve their discovery process by helping them better understand new chemicals and materials. In the financial world, they count three major financial institutions as clients and have even developed some alternative data sets that “infer trader price expectations with a sentiment model driven by options and futures positions.” These are available through the CME DataMine offering alongside data from RS Metrics and Orbital Insights.
In the insurance world, they count Allianz as both a customer and an investor. In healthcare, they’ve helped Biogen develop novel algorithms for drug discovery. In short, they’ve built a solid business that can thrive whether quantum supremacy gets here or not. And according to Mr. Downs, he’s optimistic about what’s been happening lately and introduced us to the term “quantum inspired”. As the name implies, “quantum inspired” refers to new approaches we can take to optimize algorithms that result from inspirations we encounter while researching quantum computing.
If you want to go see for yourself the sorts of things that are happening in this space, you have to understand some of the technical stuff. If you do, then wander over to the Quantum Algorithm Zoo, a comprehensive catalog of quantum algorithms. It’s the growing interest in quantum computing that excites Mr. Landon the most. He sees an ever-increasing number of research papers, researchers, quantum engineering programs, and companies, dedicating time and energy to break down the barriers towards achieving quantum advantage. Along the way, some of the quantum inspired stuff is pretty amazing.
Retail investors will want to know who is winning the race. Is it D-Wave and their newly announced 5,000-qubit device? Is it IBM? Is it Microsoft? The reality is, trying to compare these firms with each other is like comparing apples to oranges, even when looking at the few simple metrics used by us commoners to gauge progress. The first is the “number of qubits” where a bigger number is better. The second is “error rates” where a lower number is better. “Error rate” is one of the biggest impediments to achieving quantum supremacy. We need tons of qubits to do useful things, but whenever we increase the number of qubits, the error rate increases. Roughly speaking.
Quantum Apples and Oranges
You will often see companies talk about some new benchmark they set in terms of error rate and qubits. Just a few weeks ago, IBM said the following:
IBM Q System One’s performance is reflected in some of the best/lowest error rates we have ever measured. The average two qubit gate error is less than two percent, and the best gate has less than one percent error rate.
This means absolutely nothing if it can’t be scaled. Mr. Downs described simple metrics like these as the sticker price on a car which tells you very little unless you know how many miles the car has, what features it has, what model it is, and the list goes on. So how will we know when we’ve reached a “quantum advantage?” We’ll know when there’s an actual use case that’s being tackled. A very specific use case, which we will then reverse engineer to find other use cases that are similar enough to be tackled in the same manner.
There’s only one problem though. We may not know when that use case gets here because it’s highly likely that whoever develops that use case doesn’t want anyone else to know about it. It’s not some big conspiracy theory, but rather the fact that whoever gets the “quantum advantage” doesn’t necessarily want everyone else to know about it. It’s the same reason why all of 1QBit’s clients demand such high levels of confidentiality around what they’re working on. Still, Mr. Down speculates that the first viable use case may come in the area of quantum chemistry.
Building Quantum Software
When it comes to the products that 1QBit is building, they have something called 1Qloud which they wouldn’t talk about except to say that it’s “coming soon.” We already know about their software running on Fujitsu’s (also one of their investors) Digital Annealer, a special-built ASIC (a specialized piece of computer hardware) that has a circuit design inspired by quantum computing.
In other words, it’s “quantum inspired.”
1QBit’s technology has helped make the Fujitsu Digital Annealer a reality, and they’re designing their entire application layer to work with a variety of hardware. To do that, they hire experts in exotic hardware who can find out where the bottlenecks are. They’ve spent the last 5 years understanding a subset of hard industry problems which then lets them use tools like machine learning to tackle them.
Mr. Downs said that when it comes to using machine learning with special purpose hardware, they decided to spend a great deal of time working on machine learning theory. In other words, they looked at methods and approaches that differ from the mainstream. He said that using machine learning for error correction is promising, and this is something that researchers have made progress on recently. In order to build quantum computing software, you need highly-skilled developers with strong backgrounds in areas like physics and math. 1QBit scours the world looking for such people, so if you’re one of them, they have a job for you.
People from all over the world come to work at 1QBit, and this means that talent is reasonably distributed internationally. So are the research efforts being made in quantum computing. If you look at quantum research holistically – government, industry, academia – many countries like the UK, Australia, Switzerland, and China, are making good progress in various areas. It all goes back to the most exciting thing about quantum computing today – more people are working on the problems that need to get solved.
1QBIt has a viable business, whether quantum computing gets here or not, but what excites them most is the notion of being able to run algorithms on a quantum computer someday. These are algorithms that have been theoretically proven to work on a quantum computer, not just ideas.
During our conversation, Mr. Downs relayed to us an example of a top researcher at Microsoft – Matthias Troyer – who took a problem that would take 30,000 years to solve, then developed an algorithm to solve it in 30 years. Then most recently, announced that he developed an algorithm that can solve the same problem in minutes. Now that’s a quantum advantage.
Pure-play disruptive tech stocks are not only hard to find, but investing in them is risky business. That's why we created “The Nanalyze Disruptive Tech Portfolio Report,” which lists 20 disruptive tech stocks we love so much we’ve invested in them ourselves. Find out which tech stocks we love, like, and avoid in this special report, now available for all Nanalyze Premium annual subscribers.