There’s been a lot of excitement building recently about D-Wave Systems. You probably saw the Time Magazine article or perhaps the more recent Wired article on D-Wave’s progress in building a quantum computer. Both magazines take a pretty balanced look at the promise that D-Wave holds for the world and the questions that remain to be answered before that promise becomes reality. For Harris & Harris Group, Inc. (H&H), the high-profile publicity of D-Wave leads to inquiries from people who want to invest in D-Wave. The beauty of our being a publicly traded venture capital firm is that our investors are already, in a sense, early stage investors in D-Wave and quantum computing. And by investing in Harris & Harris stock (NASDAQ: TINY), new investors can get an exposure to the future of quantum computing as played by D-Wave Systems.
At H&H, we’ve been very supportive of D-Wave since the beginning and are cautiously optimistic about its long-term success. Serving on the D-Wave board of directors since the company’s early days, I am proud of what D-Wave has achieved. The jury is still out on the scaling behavior of D-Wave Adiabatic Quantum Computing (AQC) chips, yet the number of technical breakthroughs delivered by the company since we invested in D-Wave in 2005 has been staggering. From the very beginning, it was clearly an outlier deal — an incredibly risky endeavor that, if it succeeded, could fundamentally change computing and provide disproportional return to its investors.
What D-Wave was trying to do was and is incredibly complex from scientific, technological, and manufacturing standpoints. What makes such a project specifically challenging is that the probability of failure tends to grow as a square of complexity. Getting the right financial resources in place added yet another level of complexity, turning D-Wave Systems into something of a Manhattan Project. It is a huge undertaking that pushes the boundaries of science and technology, an approach that is orthogonal to the smaller, more derivative startups venture capital is funding these days.
Getting in on the ground floor of a company like D-Wave is always risky, but D-Wave made some smart moves from the very beginning, moves that impressed us as early stage investors and indicated that the company had the fortitude to stick it out for the long haul.
First, D-Wave leveraged existing manufacturing technologies and process flow. Some readers may remember that during the Cold War, the U.S. government was funding a big initiative to develop super-conducting electronics, which operated at tremendous speed at near-absolute zero temperatures. The Cold War ended, government funding dried up, and the ordinary silicon computers of the time were getting fast enough that there was really no commercial interest in super-conducting electronics anymore. D-Wave recognized that these super-conducting technologies could be used to build a quantum computer, and it was lucky enough to attract some of this expertise and hire Eric Ladizinsky, who is now the head of D-Wave’s chip manufacturing.
Second, while D-Wave started as an intellectual property (IP) play, they quickly realized that no other company but D-Wave was in the position to build a working quantum computer. The academic groups have been predominantly interested in designing and testing components for the future machines. D-wave took an engineering approach, building their system “fast and dirty,” relying on empirical studies, and moving forward with “good enough” trade-offs. There are four or five different approaches to building a quantum computer today, but D-Wave’s use of super-conducting flux qubits is the only one that currently works at scale. D-Wave was able to leverage a conventional semiconductor foundry to build its chips, which is critical for any mass-produced electronic component. Using these existing technologies keeps costs down, accelerates the research, and enables D-Wave to iterate on new designs quickly. These are strategies that appeal to early investors such as H&H and reduce the overall risk of our investment as well as its capital intensity.
There’s no doubt in my mind that what D-Wave has built is quantum. The chip demonstrates both local entanglement and tunneling, phenomena that are not available to us in the classical regime. The big question, however, is, “Does it matter?” D-Wave still needs to show that what it has built scales faster relative to problem complexity than a classical computer.
It’s a funny thing, but to simulate one computer chip, you need another computer chip to do simulations and make predictions. To simulate a quantum computer, you really need another quantum computer – the company is well beyond what can be simulated by normal digital machines. There is really no known way to predict the performance of D-Wave’s next generation chip. We simply have to build this chip and measure how it performs. This approach is rather unorthodox in the computer science and mathematical community today, but many breakthroughs in human knowledge have been based on empirical observations that eventually fed into theoretical models leading to deep understanding of the discovered phenomena. It may not sound utterly elegant, but D-Wave decided to build its future machine using empirical observations instead of over-theorizing about its properties and performance limitations.
For now, D-Wave is the undisputable leader of scalable quantum computing. With over 100 U.S. patents issued, the company has a firm grip on the intellectual property estate around quantum computing, especially in superconducting realization of the hardware platform. It fully controls its proprietary manufacturing flow, EDA tools, chip-design, and system architecture. Even the dilution fridges they use are not off-the shelf; they need to undergo massive upgrades at the company before they can be deployed to the field. By the nature of its machine, the D-Wave platform should be relatively immune to the competition from traditional digital computers, which translates into high margins for its unique capabilities.
Since our original investment back in 2005, I have been consistently impressed by D-Wave’s ability to execute on aggressive technical schedules. The company designed and fabricated the first quantum chip, and it has become a catalyst in multiple areas of superconducting electronics and cryogenic equipment. We remain optimistic about D-Wave’s ability to execute in the future and to turn quantum-computing into reality. Today D-Wave’s machine may be suitable only to visionary and well-funded organizations facing critical complex problems, such as Google, Lockheed Martin, NASA and other government agencies. Tomorrow, with the right amount of funding and execution, D-Wave solutions may appeal to a much wider audience. D-Wave is a well-funded group with a Tier-1 management team that is selective about its investor base. However, there is still time for investors to join Harris & Harris Group and get in on the ground floor of quantum computing, however small the allocation might be.