In his opening keynote at D-Wave’s annual Qubits 2024 user meeting, being held in Boston, yesterday and today, CEO Alan Baratz again made the compelling pitch that D-Wave’s brand of analog quantum computing (quantum annealing) is the only quantum technology delivering commercial value today. Many would agree.
- D-Wave’s customer base is expanding and at least one customer that has rolled out a second application to production following a successful first deployment.
- Progress on Advantage2, planned to have 7000 qubits, is proceeding on schedule.
- D-Wave continues to roll out improved tools and hybrid solvers that ease application development and deployment.
- It has a paper claiming quantum supremacy that’s in peer review and is expected to be published. (There’s a pre-press version.)
- D-Wave’s gate-based development program, which leverages lessons learned from annealing work, is making progress and D-Wave plans to join the wave of companies offering gate-systems when they are ready – something it doesn’t expect soon.
“We are at a watershed moment for quantum computing. We are at the point where quantum computing is transitioning from research experimentation to real business applications in production helping companies to improve their business operations. But there’s only one company in the world that can say that, and that’s D-Wave Quantum,” said Baratz, “And the reason why we can say that is we’re the only company in the world that has quantum computers that are large enough and powerful enough to actually be able to support those applications, and we are just now at the point where we are seeing customers actually transition applications into production use.”
In practical terms, D-Wave is walking its talk as demonstrated with new collaborations, such as one with Los Alamos National Labs in materials science, new commercial “sales” such as plans announced yesterday by defense contractor Davidson Technology to put a D-Wave system at its Huntsville AL HQ, an expanding international footprint (e.g. Quantum Basel), and solidifying consensus that D-Wave’s quantum annealing approach (and analog quantum computing generally) works well on optimization problems. Also, the company’s stock (NYSE QBTS) seems relatively stable and recently became part of the Russell 3000 Index.
On the one hand, D-Wave is energized by progress and ambition with chief cheerleader and chief rival critic Baratz leading the charge. On the other hand, Baratz’ dogged Rodney Dangerfield-like mantra that ‘D-Wave gets no respect’ mantra is wearing thin. Maybe that’s the nature of the fight.
The war of words over quantum’s slippery vernacular — the true meaning of quantum advantage, quantum supremacy, and quantum utility — and trash-talking competitors (if not by name) continues and probably wasn’t necessary. Maybe it’s mostly a flare to draw attention anyway. One has the sense that Baratz grits his teeth every time he hears a certain three letters of the alphabet and maybe a company named alphabet as well. Time to get over it?
The quantum computing landscape still defies easy description and prediction, but if there is indeed a NISQ (near-term intermediate scale quantum) window for success, D-Wave seems well positioned to grab it. While Baratz unabashedly slammed historic and current criticism of D-Wave Technology, which is his usual practice at these events, he also offered something close to a mild mea culpa in Q&A. Asked how potential users can make sense of the flood of papers, often conflicting, pouring out into the community, he said:
“First of all, we at D-Wave went through a what I’ll characterize is a dark period, I don’t know, maybe seven or eight years ago, [and] we ended up making some statements that were, frankly overstatements, and we got slammed very, very hard for doing that. We were just way out in front of our skis, and from that point on, said that we would always be careful, thoughtful, scientifically grounded, measured in everything we do. We would not make a statement or put out a result unless we had confidence that it was scientifically sound and could be backed up.
“Unfortunately, I think, in their desire to demonstrate progress, we’re seeing way too many companies in the quantum space now get out in front of their skis. And unfortunately, since they’re all kind of in it together. They’re not checking one another, and so we end up seeing way too many results and papers that are just overstatements of reality, and we need to stop letting that happen. We need to monitor that. We need to police, that we do it when we see a paper that one of our customers has written on a result using our system, if we can’t repeat it, if we can’t justify it, we’ll say, no, do not talk about this. This is not a solid result. We just need to hold the industry accountable.”
OK, not exactly a mea culpa. Nevertheless, few would disagree that sorting wheat from chaff in the young but high-stakes quantum computing market is difficult at best. For all the time taken to dissect what quantum advantage, quantum supremacy, and quantum utility really mean, or should mean, it’s not really clear that’s a worthwhile effort at this point. The potential user community really doesn’t care – a point made well on a customer panel during the conference about which HPCwire will have coverage later.
They care about applications and ROI versus classical systems (or any other systems).
There wasn’t much breaking ‘news’ but more of a summing up of D-Wave’s plentiful progress since its last Qubits meeting. Much of the material was familiar from its recent analyst day (January 2024). Much of the meat of the conference will come on day-two when sessions will drill down for the user community. (Link to video of keynote. D-Wave will post other recorded sessions next week.)
On the technology front, D-Wave reported steady progress. Its Advantage2 system is expected in 2025/26 time frame; last year, D-Wave launched an 400-qubit prototype, and last quarter introduced a 1200-qubit prototype, each featuring advances. The Advantage2 system will have longer coherence times, 20-way connectivity, and increased energy. In April, D-Wave introduced Fast Anneal protocol on its Leap real-time quantum cloud service.
Baratz said, “We recently launched a new protocol for controlling the quantum computer. It’s called fast anneal. Essentially, what it means is that we can run our annealing algorithm much faster than we were able to run it previously, and, in fact, run it so that most of the time the anneal is running, we’re in the coherent state with the quantum processor.
“The combination of the Advantage2 system and Fast Anneal was what allowed us to demonstrate in Nature 2023 last March, a year ago, that the system is actually performing coherent quantum annealing, and that when the system is performing coherent quantum annealing, we get a scaling polynomial speed up over classical on spin glass optimization problems. It’s also the fast anneal and the 1200 qubit advantage quantum computer that is what has allowed us to achieve the quantum supremacy result, and this is now available in leap for anybody to use as well.”
LANL collaborator Carleton Coffrin, a presenter at the conference, said “When we saw the Advantage2 [announcement] come out, the claim was that it was have higher energy scale, lower noise and better coherence time. Using the protocols we’ve developed over the years for benchmarking these systems, we were able to quickly verify the higher energy scale and a lower noise property. It is difficult to verify the higher coherence time without the Fast Anneal but we’re extremely excited about the release of that feature.
“Since the first paper came out in 2022 we’ve kind of been chomping at the bit to play with it, and now we’re super excited that we have the time to do that. The theoretical physicists at LANL have kind of a huge list of ideas, and it’s going to take us a while to work through them and really do some cool science,” said Coffrin.
Customer stories are always a core part of these meeting and D-Wave had several onhand and presenting. GenAI was also significant theme, with presenters from ZapataAI and long-time D-Wave collaborator SavantX tackling the broad topic of blended GenAI-Quantum solutions. You may recall that SavantX developed the Port of Los Angeles application which uses D-Wave’s quantum computer. The second half of HPCwire’s article, ‘Eyes on the Quantum Prize – D-Wave Says its Time is Now’, has a fair amount of material on how D-Wave engages customers.)
Collaborators and customers who took the stage with Baratz included:
- ZapataAI (Jon Zorio, CFO) — D-Wave announced the collaboration with ZapataAI in February, “to develop and bring to market commercial applications that combine the power of generative AI and quantum computing technologies.”
- Artificial Brain (Dana Linnet, Quantum Market Engagement) — defense an aerospace quantum software specialist. Linnet discussed two applications, including satellite tracking, noting this is an optimization problem that’s difficult for classical systems but well suited for D-Wave.
- QuantumBasel (Damir Bogdan, CTO) — An ongoing collaboration which includes providing access to a D-Wave Advantage through QB to Europe-based users.
- Pattison Food (Lindsay Dukowski, senior analytics leader) — Patterson rolled out a scheduling application a year or so go and has added a second.,
- Davidson Technologies (Dale Moore, president) — Defense contractor Davidson discussed an application developed for missile defense developed with D-Wave.
- QuantumBasel (Damir Bogdan, CTO) — An ongoing collaboration which includes providing access to a D-Wave Advantage through QB to Europe-based users.
- LANL (Carleton Coffrin, staff scientist) — Unusual magnetism is one of the key focus areas LANL plans to tackle.
It’s best to track down the details of these engagements separately and D-Wave is happy to provide them. Most are optimizing programs. Perhaps most impressive were Dukowski’s comments on building and deploying a second quantum application. The first application was an ecommerce driver auto scheduling application. The second is for auto scheduling in our stores.
“One of our largest costs is our labor in our stores. So we’ve been trying, really, for decades, to try to buy a system that could do auto scheduling for us, because that’s one of our most manual processes. So we’ve been looking at top vendors in the space for Workforce Management and scheduling, and they’ve promised they could do it,” said Dukowski. “When we brought them in, they admitted, no, sorry, our environment’s pretty complicated. We have over 13 different collective bargaining agreements, unionized environments. The scheduling rules are just too complex to be solved with classical compute.
“The second application we did on our own. It was relatively easy. I have a very talented team. We learned a lot. We had a lot of support with D-Wave, kind of exploring how we could leverage the technology to solve our problems.”
Without doubt, D-Wave has faced critics throughout its journey. Clad in a smoking jacket, a top-shelf-beverage nearby, Baratz closed his keynote reading the book of D-Wave, as it were, likening the 25-year race as a tortoise and hare parable.
“This is a story of D-Wave’s steady race to a superposition. Once upon a time in a beautiful place where the sea meets the mountains, a quantum computing company was born. D-Wave, from the day it came into existence more than 25 years ago, has had a clear mission to harness the incredible power of quantum technology and put it to use in solving our most challenging problems. It’s not the largest or the loudest character, but rather brings a customer focused scientifically validated and ever measured approach to building successful quantum businesses, much like the Aesop Fable the Tortoise and the Hare,” read Baratz.
“D-Wave, the tortoise hasn’t always been recognized for its ability to fiercely compete. This is a story of vision, resilience, conflict, execution, and ultimately, success, all the great things that a story needs to have. As with many of our favorite stories, this one is also filled with villains who are eager to take down those they fear, so on its journey to becoming the world’s first commercial quantum computing company, D wave has had to fight many battles, slay many dragons and protect the voices of those seeing great outcomes with its solutions.”
It was tempting to use the photo above and this part of the Baratz keynote as the introduction to this article. Drama. But even if (still) true, after all this time, the relentless retelling of the story seems less productive now – we get it! – and undermining to the steady one-foot-in-front-of-the-other painstaking progress that has brought D-Wave to the edge of success, at least in the NISQ era. The battle is hardly over.
D-Wave is not the only game in town, although for moving quantum computing into or at least towards production environments, it may well be the farthest along. Neutral atom-based quantum systems, which also currently use an analog approach to QC, are gaining attention and show strength in optimization applications, which Baratz calls the low hanging fruit in QC. Gate-based development by IBM, Google, IonQ, Quantinuum, et al, certainly aren’t slowing down but seem further from meaningful revenue goals.
Stay tuned.
Link to Qubits 2024, https://www.qubits.com/