D-Wave Systems Inc., is the only company in the world to build both quantum annealing and gate-based quantum computers. In an interview, the company's vice-president of product management Murray Thom explains why the company is finally building gate-based quantum computers and shares his thoughts on when the world may see quantum supremacy. He also talks about how India fares in the quantum computing space. Thom has led teams engaged in customer projects related to algorithms, applications, and performance testing, and even assembled a few early quantum computers by hand. In his earlier roles at D-Wave, he was responsible for the development and delivery of the Leap quantum cloud service and the Python-based Ocean open-source tools. Edited excerpts from the interview:
How much progress has D-Wave made with its Leap quantum cloud service? Can you share some adoption numbers for India?
We launched Leap in October 2018, but provided access to users in India in 2020. We already have over 25,000 sign-ups making India the third-highest country in terms of LEAP sign-ups since its launch. And we have seen that the number of people accessing quantum computers in the cloud through Leap from India has increased by 58% since January 2021. India is a leading country in terms of software development expertise. It has an appetite for advanced technologies, and Leap is all about providing people with access so that they can learn the technology, the programming models, and how to formulate problems for the system. We've provided open-source examples to allow them to do that. We want to follow the lead of the users there in terms of their ability to show progress with practical applications and building enterprise solutions.
You also have a partnership with the University of Southern California that has been working with a D-Wave quantum computer.
Yes. The University of Southern California built a quantum computing center in 2011. They have been able to leverage our quantum computer to do some real cutting-edge research stuff in the quantum computing space. In terms of practical applications, they have published some great work including how machine learning can work with our quantum computing systems to advance data analysis for the discovery of Higgs bosons; and methods for doing error correction with quantum annealing.
With the upgrade to D-Wave’s Advantage quantum system,--our first 5,000+ qubit machine physically located in the US at USC’s Information Sciences Institute--the university and D-Wave will increase the capacity for academic researchers, government users, and the business community to continue studying how quantum effects may speed up the solution of complex optimization, machine learning and sampling problems, and new breakthrough results in quantum optimization. The Advantage system (Advantage2 is expected to feature 7,000 qubits with a new qubit design) is accessible via the Leap quantum cloud service.
What explains your adopting the gate model after bettering on the quantum annealing model till date?
According to the Boston Consulting Group, quantum computing can create a value of $450 billion to $850 billion in the next 15-30 years. Some of that is in the space of quantum chemistry simulations, and differential equations. Another portion of it is in optimization. If we want to be able to address the quantum chemistry simulations, we need to be able to build these gate model quantum computers, and D-Wave wants to be the provider of everyone's full quantum solution, which is why we're expanding our product line to include those gate model systems. The other important thing to note is that using gate model machines to do optimization is incredibly inefficient. This means we need both (quantum annealing and the gate model) for our customers in the future.
What's the D-Wave strategy to simplify quantum computing for enterprises?
D-Wave is completely focused on showing the business value in practical quantum applications. We believe this is the right strategy towards building a new disruptive technology. We're building two of the most commonly known models for quantum computing -- the gate model for quantum computing and the annealing model. And we've had a lot of traction. We're finding that our customers are able to see value in their applications right away.
As an example, CaixaBank is applying quantum computing in investment hedging in the insurance sector. They used our quantum hybrid solver services (use of both classical and quantum resources to solve problems) to code a faster algorithm to reduce the computing time needed to reach an optimal solution to improve investment portfolio hedging by up to 90% over the traditional solution.
Likewise, we have worked with Save-On-Foods--a mid-sized grocery store chain in Western Canada. They used our hybrid quantum algorithms to reduce the time for grocery optimization from 25 hours down to just two minutes. Another example is that of Menten AI which has been able to quantum design proteins, synthesize them, and take them for live virus testing of Covid-19. Besides, with the open-source Python Software Development Kit, it's (quantum computing) actually very accessible for people to work on problem formulations, and see that practical value quickly.
When do you see the world getting a stable quantum computer?
Your users might be surprised to hear this but quantum computing in the cloud is stable right now — we've got 99% uptime for our quantum computers. You can program the system and get responses back in seconds. There's so much mystery and complexity about quantum computing -- it's a blessing and a curse, and we have to live with that complexity and make it simple for people to see how to program it.
Our 5000-qubit quantum computer is the largest programmable quantum computer that's available today. There's a technology stack of powerful algorithms that people are developing to leverage it in these hybrid contexts. There are already cases where people are basically able to make their businesses more efficient, more responsive to these kinds of supply chain disruptions to allow us to basically adapt and become efficient in order to keep our operations functioning using quantum computing technology. We have really realized we need to focus on practical quantum computing — the whole industry needs to focus on that. I think users and enterprises and CIOs are really going to find that that helps them.
But do we have enough skilled programmers for quantum computing?
You don't have to be a quantum physicist to program quantum computers. It's our job to bring programming tools that are familiar, like Python programming kits with accelerated C++ algorithms. We have to do more in terms of making the mathematical aspects much more digestible and easier to see. That's why we're focused on these open-source examples such as the n queens (problem of placing n queens on an n x n chessboard such that no two queens attack each other) example. That's not because a lot of folks are working on chess examples in the industry. Rather, when you're programming a calendar for a workforce, you need to know how am I going to use variables to describe the days of the week or the hours in the day, and the techniques that are shown in those test examples on how to choose those parameters.
Advanced technologies like artificial intelligence (AI), Internet of Things (IoT), etc., appear to be leveraging quantum computing.
The cloud was the best thing that ever happened to quantum computing (D-Wave's Leap is cloud-based). IoT is allowing people to operate their business more efficiently with sensors, but they (businesses) also need more advanced planning and optimization techniques to be able to leverage that effectively. AI also needs advanced compute tools, and there's a very close connection between the underlying instructions that quantum computers are working on the optimization problems and the types of models that are used in AI and Machine Learning. There's a lot of complementary work that's happening here — sometimes at the technology level, and sometimes at the application and business level to create solutions. These are exciting times.
How do you address concerns that quantum computing will become as formidable as AI and impact us?
That's a good point. It often makes me think of Isaac Asimov who, when writing about robots, acknowledged that new knowledge has its dangers but asked if the solution is to retreat from knowledge, or make it possible to use the knowledge to advance what we're doing to take its benefits and mitigate its downsides. That was such a smart way of interpreting it. I mean, it's all on all of us to use what we're learning about these new technologies, and with quantum computing, to bring the positives of disruption and figure out how to exclude the negatives. A great example of that is these hybrid solvers that leverage the best of what classical computing and quantum computing can offer.
When do we see quantum supremacy happening?
People define quantum supremacy in different ways. There's a lot of focus on whether we do a calculation that nobody could ever do with classical computing -- that will be an important milestone. But what we're focused on right now is how businesses can use this (quantum computing) to better what they're doing today. We're focused much more closely on that commercial advantage, and we're already seeing it.