Title

Presentation

Operator: Good morning ladies and gentlemen and welcome to the D-Wave Q3 2024 earnings conference call. At this time, all lines are in listen-only mode. Following the presentation, we will conduct a question and answer session. If at any time during the call you require immediate assistance, please press star, zero for the Operator. This call is being recorded on Thursday, November 14, 2024. I would now like to turn the conference over to Kevin Hunt, Investor Relations. Please go ahead.

Kevin Hunt: Thank you and good morning. With me today are Dr. Alan Baratz, our Chief Executive Officer, and John Markovich, our Chief Financial Officer. Before we begin, I would like to remind everyone that this call may contain forward-looking statements and should be considered in conjunction with cautionary statements contained in our earnings release and the company’s most recent periodic SEC report. During today’s call, management will provide certain information that will constitute non-GAAP financial and operational measures under SEC rules, such as non-GAAP gross profit, non-GAAP gross margin, non-GAAP adjusted operating expenses, adjusted EBITDA loss, and bookings. Reconciliations to GAAP financial measures and certain additional information are also included in today’s earnings release, which is available in the Investor Relations section of our company website at www.dwavequantum.com. I will now hand over the call to Alan.

Alan Baratz: Thanks Kevin, and good morning everyone. It’s great to be with you again, and thank you all for joining us today. It’s an exciting time in the quantum industry with interest levels increasing at a rapid rate, and I’m pleased to be able to update you on the momentum that we are seeing here at D-Wave. Interest in our technology continues to grow and we believe it is evident that annealing quantum computing is the critical accelerant to commercial adoption of quantum technology, and let me be clear – it is our 5,000 qubit Advantage quantum computer, the largest and most powerful quantum computer in the world, that is driving this adoption. Moreover, it is showing up in bigger wins like NTT DOCOMO, Japan’s largest wireless carrier with more than 87 million subscribers, where our quantum application is slated to move into production.

I’d like to start today’s business update by highlighting some product and technology advancements. We recently announced that we have completed the calibration of 4,400 qubit Advantage 2 processor which represents a significant step towards the commercial release of Advantage 2, which is our sixth generation annealing quantum computer. The 4,400 qubit processor provides impressive performance gains over the existing Advantage system, including double coherence time for faster time to solution, a 40% increase in energy scale for higher quality solutions, and increased qubit connectivity from 15 to 20-way to enable solutions to larger problems. With these substantial performance improvements, we believe that the Advantage 2 will be able to solve our customers’ increasingly complex computational problems, especially in areas such as optimization, AI, and material science.

The industry is taking note of this remarkable performance. Heather West, quantum analyst at IBC, a premier global market intelligence firm, recently published a very positive report on Advantage 2, called D-Wave: Setting New Standards in Quantum Computing Performance. In that piece, she said – and I’m quoting from the piece, to date D-Wave’s quantum annealing systems are one of the few quantum computing technologies used by organizations to run production use cases to gain business value. The range of optimization problems which can be applied to the annealing quantum computer demonstrate a versatility once thought to be limited to gate-based systems. This versatility will help to expand the adoption of quantum computing to a larger number of potential end users across multiple industries; consequently, it may be that the long sought after killer use case for quantum is in fact quantum optimization.

That’s the quote from Heather. I want to add that there’s also a large body of evidence that only annealing quantum computing systems will actually be able to deliver speed-up on quantum optimization problems, what Heather is calling the killer use case for quantum. As a reminder, we also have an extensive and valuable quantum patent portfolio. This includes over 240 U.S. granted patents with the majority covering both annealing and gate model technologies; for example, we own fundamental IP around on-chip control and read-out for quantum processors that support addressing and data pipelining on the quantum processor chip. This overcomes a fundamental challenge related to controlling large quantum processors without the need for large numbers of complex and costly IO lines.

We are using this technology in our annealing quantum computers today and it will be required in scale gate model systems of the future. We continue to make rapid progress on exploring generative AI architectures that directly use quantum processing unit QPU samples from quantum distributions to facilitate faster and more energy efficient model training and inference. We have completed the initial designs of transformer and diffusion architectures and are moving towards benchmarking the role of the QPU. We are also directly supporting several customers investigating restricted Boltzmann machine architectures, a canonical machine learning approach for generative AI that uses samples from the QPU. On the software side of the business, we introduced service level agreements – SLAs for Leap quantum cloud service customers who are transitioning applications into production.

We believe D-Wave is the only quantum computing company providing formal SLAs. It’s a significant move that reflects Leap’s high levels of availability, reliability and performance, underpinning its ability to support commercial grade applications as customers move into production deployments. In fact, our monitoring data shows that the Leap service has consistently exceeded 99.9% availability over the past two years, meaning that the service is highly reliable even during periods of high demand. Since the launch of the Leap service, customers have run nearly 2 million jobs across our three production Advantage systems without having to schedule work in advance, endure lengthy queue times, or work around unavailable hardware. All of those things seem to be the norm for other quantum cloud services.

Beyond the introduction of SLAs, we also improved the performance of our new, non-linear hybrid quantum solver capable of supporting optimization problems with up to 2 million variables and constraints. For example, on an important customer problem related to capacitated vehicle routing, the problem can now be solved in one-fifth of the time. These technology advancements are driving real business solutions. This quarter, we announced a major quantum optimization pilot with NTT DOCOMO in Japan. The pilot was held across three geographic regions in Japan and delivered a 15% reduction in network congestion, and the D-Wave solver completed the required tasks in 40 seconds compared to the 27 hours required for classical compute methods. Upon completion of the pilots, NTT DOCOMO plans to move the annealing quantum computing solution into production across Japan.

This will be a major in-production application running on D-Wave technology, and as we’ve seen with other customers after a successful first application, NTT DOCOMO is now exploring additional opportunities to deploy annealing across other business units. We believe that we are the only quantum computing company in the world that has quantum applications being used in commercial production deployments. I’d also like to highlight a new AI-related collaboration with Japan Tobacco, where we are building a proof of concept that leverages both quantum computing and AI in the drug discovery process. The company is seeking to create a new process for discovering first-in-class pharmaceutical small compounds, and our annealing solutions show great promise.

This past quarter, we also worked with customers on a variety of other hybrid quantum applications, including optimization of manufacturing lines, financial portfolio optimization, optimization of delivery truck routes and scheduling, optimization of agricultural resources and operations, and many others. On the government side of the business, I’m delighted to say that D-Wave has been deemed awardable on the U.S. Department of Defense’s Tradewinds buying platform, which is designed to accelerate the procurement and adoption of emerging technologies. The designation means that D-Wave’s annealing quantum computing products and services are now available on the marketplace alongside other offerings, like AIML data and analytics capabilities.

We also saw continued interest from national security and civilian agencies around near term quantum applications, highlighted by a U.S. Department of Defense RFI asking for near term application identification as well as increased discussions relating to optimization of emergency response, transportation, and other public sector services in the U.S. and abroad. This increased activity showcases government’s interest in better understanding how annealing quantum computing can play a pivotal role in solving public sector problems to date. Broadening our reach, in September we announced a partnership with Staque, a leading technology consultant in the Middle East, where we will jointly help customers develop and deploy hybrid quantum applications designed to address enterprise optimization and AI problems.

We announced this partnership at the first-ever Qubit UAE, a half-day version of our U.S.-based event, where we showcased a variety of use cases utilizing quantum; and by the way, our next annual Qubit user conference will take place at the end of March in Scottsdale, Arizona, where we will once again showcase the latest D-Wave Quantum technology innovations alongside remarkable stories of customer success spanning quantum optimization, quantum AI, and quantum research. We invite you all to attend and registration is now open. In September, we also announced that D-Wave has joined the Chicago Quantum Exchange as a corporate partner. As the only commercial quantum company, we share CQE’s vision of developing practical optimization use cases.

We aren’t just building cutting edge technology at D-Wave, we’re building a company for the long term. In September, we announced that Sophie Aims has joined D-Wave as Chief Human Resources Officer, where she will lead our human capital innovation strategy essential to supporting our go-to-market growth objectives. Sophie brings over 25 years of experience at a variety of public and private international technology and services companies. In October, we announced the appointment of two tech industry veterans to our board of directors. John DiLullo, CEO at Deepwatch, brings over 30 years of experience at leading technology firms; and Rohit Ghai, CEO of RSA, also brings a wealth of experience from both technology start-ups and large enterprises across software, services and security.

These are all leaders with an aggressive growth mindset and experience commercializing new technologies. Finally, I will close with a quick comment on our financials. Subsequent to the end of the quarter, we completely paid off our $50 million secured term loan with PSP and we have nearly $30 million in cash. With that, I will hand the call over to John to provide a review of our third quarter and first nine months of 2024 results. John?

John Markovich: Thank you Alan, and thank you to everyone taking the time to participate in today’s call. In my review of the third quarter and nine months results, I will be providing non-GAAP operating metrics that include bookings, as well as non-GAAP financial metrics that include non-GAAP gross profit, non-GAAP gross margins, and adjusted EBITDA loss, as we believe these metrics improve investors’ ability to evaluate our underlying performance. These measures are defined in the tables at the bottom of today’s third quarter earnings press release, with the non-GAAP financial metrics for the most part adjusting for non-cash and non-recurring expenses. Revenue in the third quarter of fiscal 2024 totaled $1.9 million, a decrease of approximately $700,000 or 27% from the third quarter of fiscal ’23 revenue of $2.6 million.

QCaaS revenue for the third quarter was $1.6 million, an increase of $500,000 or 41% from the fiscal ’23 third quarter QCaaS revenue of $1.1 million, with the increase principally due to an average revenue per customer–a higher average revenue per customer for D-Wave’s QCaaS subscription services. As a result of the timing of closing new professional services agreements, professional services revenue for the third quarter was $300,000, a decrease of $1 million or 80% from the fiscal ’23 third quarter professional services revenue of $1.3 million. Although the third quarter professional services revenue declined on a year-over-year basis, the third quarter professional services bookings were quite strong, representing the second highest quarterly bookings quarter over the last three years.

We expect the fourth quarter revenue to improve over the third quarter revenue. Bookings for the third quarter totaled $2.3 million, a decrease of $600,000 or 22% when compared to the third quarter of 2023 bookings of $2.9 million. Again, we expect the fourth quarter bookings to improve over the third quarter bookings. With respect to the up-leveling of our go-to-market organization that we highlighted in our last earnings call, we have made significant progress over the last several months. During this period, we have filled 11 positions across the United States, Germany, Switzerland, and the U.K., including four quota-carrying positions and several positions focused on the manufacturing, logistics and services verticals. We expect to substantially complete the up-leveling of the organization by the middle part of next quarter.

With respect to customers, we continue to broaden and diversify our customer base across commercial, government and research customers. In comparing the most recent four quarters with the immediately preceding four quarters, D-Wave had a total of 132 customers compared to a total of 125 customers, 76 commercial customers compared with 75 commercial customers, with the commercial customers including 27 Forbes Global 2000 customers compared with 26 in the prior period. Our Forbes Global 2000 customers constituted 36% of the total number of our commercial customers. Our Forbes Global 2000 customers include numerous internationally based companies, including MasterCard, Ford Otosan, Vinci Energies, NEC, DENSO, Bridgestone, and Koç Holding. In comparing the most recent four quarters with the immediately preceding four quarters, revenue from government customers increased by $800,000 or 66%, revenue from research customers increased by $600,000 or 47%, revenue from commercial customers decreased slightly by $200,000 or 4% with the commercial revenue as a percentage of total revenue declining from 70% to 59%, primarily due to the company’s increased focus on government and research customers.

With respect to gross profit, the GAAP gross profit for the third quarter was $1 million, a decrease of $500,000 or 32% from the fiscal 2023 third quarter gross profit of $1.5 million, with the decrease due primarily to lower professional services revenue that was partially offset by higher QCaaS revenue. With respect to the non-GAAP gross profit for the third quarter, it totaled $1.3 million, a decrease of $600,000 or 35% from the fiscal 2023 third quarter non-GAAP gross profit of $1.9 million. The difference between GAAP and non-GAAP gross profit is limited to non-cash stock-based compensation and depreciation and amortization expenses that are excluded from the non-GAAP gross profit. With respect to margins, the GAAP gross margin for the third quarter was 55.8%, a decrease of 3.9% from the fiscal 2023 third quarter GAAP gross margin of 59.7%.

The non-GAAP gross margin for the third quarter was 67.2%, a decrease of 8.4% from the fiscal 2023 third quarter non-GAAP gross margin of 76.6%. Again, the difference between the GAAP and the non-GAAP gross margin numbers is limited to stock-based compensation and depreciation and amortization expenses. Net loss for the third quarter was $22.7 million or $0.11 per share compared with a net loss of $16.1 million or $0.12 per share in the fiscal 2023 third quarter. The adjusted EBITDA loss for the third quarter was $13.8 million, an increase of $2.2 million or 19% compared with the adjusted EBITDA loss of $11.6 million in the fiscal 2023 third quarter. The increase was primarily due to lower revenue and higher operating expenses that were principally related to the company’s increased investment in its go-to-market organization.

I will now address D-Wave’s operating performance for the first three quarters of fiscal ’24. Revenue for the nine months ended September 30, 2024 was $6.5 million, an increase of $600,000 or 11% from revenue of $5.9 million for the nine months ended September 30, 2023. QCaaS revenue for the nine months ended in September was $5.1 million, an increase of $1.8 million or 52% from the QCaaS revenue of $3.3 million for the nine months ended September 30 of 2023, with the increase principally due to higher average revenue per customer. As a result of the timing of closing new professional services engagements, professional services revenue for the nine months ended September 30 was $1.3 million, a decrease of about $900,000 or 42% from professional services revenue of $2.2 million for the nine months ended September 30, 2023.

Bookings for the nine months ended September were $5.6 million, a decrease of $1.8 million or 24% from bookings of $7.4 million for the nine months ended September 30, 2023, with the decrease entirely due to one customer terminating its operations. With respect to the GAAP gross profit, it totaled $4.1 million for the first nine months, an increase of $1.4 million or 54% from $2.7 million of GAAP gross profit for the nine months ended September 30, 2023, with the increase due primarily to the growth in revenue and lower stock-based compensation expenses in cost of sales. The non-GAAP gross profit for the first nine months was $4.7 million, an increase of $900,000 or 25% from the non-GAAP gross profit of $3.8 million for the nine months ended September 30, 2023.

GAAP gross margin for the first nine months was 62.7%, an increase of 17.3% from the 45.4% GAAP gross margin in the prior nine months, with the increase due primarily to an increase in revenue and decrease in stock-based compensation costs and increased operating efficiencies. Non-GAAP gross margin for the first nine months was 72.7%, an increase of 8.1% from the 64.6% non-GAAP gross margin for the nine months ended September 30, 2023. The net loss for the first nine months was $57.8 million or $0.32 per share compared with a net loss of $66.7 million or $0.51 per share in the fiscal 2023 nine-month period. Adjusted EBITDA loss for the first nine months was $40.6 million, a decrease of $2.8 million or 6% from the adjusted EBITDA loss of $43.4 million for the nine months ended September 30, 2023, with the decrease due primarily to higher gross profit and lower general and administrative operating expenses.

Now I will address the balance sheet and liquidity. As of September 30, D-Wave’s consolidated cash balance totaled $29.3 million, and as Alan highlighted previously, subsequent to the end of the quarter, we paid off the remaining balance of the PSP $50 million secured term loan. With respect to our capital raising capacity, on April 12 of this year we filed $175 million shelf registration statement on the Form S-3 that went effective in May, and we carved out $100 million of that for an at-the-market program. As of the end of September, we had $79.1 million of issuance capacity under the ATM program. With respect to our equity line of credit with Lincoln Park Capital, as of the end of the quarter, we still had almost $50 million, or $49.9 million remaining in available issuance capacity.

We are reiterating our fiscal 2024 adjusted EBITDA loss guidance of less than $54.3 million. To conclude, as we have previously stated, we believe that D-Wave has the opportunity to be the first independent publicly held quantum computing company to achieve sustained profitability and to achieve this milestone with substantially less funding than required by other independent publicly held quantum computing companies. With that, we will now proceed to questions.

Operator: Thank you. Ladies and gentlemen, we will now begin the question and answer session. [Operator instructions] Your first question comes from Quinn Bolton with Needham. Your line is now open.

Quinn Bolton: Hey guys, congratulations on the continued progress with customers. I guess I wanted to start with that – last quarter, you did talk about the NTT DOCOMO win and the pilot. It looks like they’re now looking to move that into production across all of the cell sites in Japan. Can you give us some sort sense, you know, what is the timing when they’ll start to move that application into production, and I think in the past, you’ve talked about production applications could generate somewhere between, say, half a million and a million of QCaaS revenue a year. Is that the right range to be thinking about for this contract, or should we be thinking about a different range?

Alan Baratz: Quinn, first of all, you’re correct that as the applications transition from proof of concept or trials into production, they generate significantly more revenue, in particular QCaaS revenue for the company, and we in fact have indicated that as applications move into production, they tend to generate on the order of high hundreds of thousands or more per year of recurring QCaaS revenue. There’s nothing that’s changed relative to the model and our expectations on the magnitude of revenue as applications move into production. With respect to the timing for NTT DOCOMO, they’ve completed the trial work across three regions in Japan and are now putting in place the plans to move this throughout the entire country, but they have not yet disclosed or made public the plans and the timeline for that rollout.

Quinn Bolton: Understood. Thanks Alan. Then interesting to see–I think it’s the first time you guys are starting to break out revenue for U.S. government, which was up. I know you guys have sort of qualitatively spoken about increasing opportunities on the government front, so maybe just expand on what you’re seeing on the government side of the business. It looks like you have some momentum there. Then perhaps any thoughts on when we might see a re-authorization of the National Quantum Initiative, and if we get that at what looks to be pretty significant increased funding levels, how would you think that could affect your business? Thanks.

Alan Baratz: Yes, so let me take the National Quantum Initiative first. As you may recall, but maybe not, the NQI was actually originally passed under Trump, and it was supposed to be renewed about a year ago at the five-year mark, and it has not been renewed. We are hopeful that with Trump back in office, he’ll focus on completing the work that he began around the National Quantum Initiative and it will help to push Congress to get that reauthorized. Once that is reauthorized, that will open up additional government funding for a variety of different quantum programs that we will benefit, not just us, but all players in the quantum industry. With respect to our progress with not just the U.S. government but governments around the world, in the U.S., being deemed awardable on the Tradewinds program is actually a very significant deal.

It’s not an easy process to be deemed awardable, but once you are, then in fact the government–DoD agencies can buy from you without needing to go through the long complex processes that have traditionally been required to sell into DoD. As the government is becoming more interested in near term quantum applications in part as a result of the language that was in the National Defense Authorization Act that got passed about a year ago, calling out near term quantum applications, we in fact are seeing interest, if not early deals actually being done with respect to near term quantum applications, and being awardable on the Tradewinds platform will make it even easier for those agencies to be able to buy from us as they focus on trying to incorporate quantum into those application areas.

We think that’s a very significant advance for us, and of course we are continuing to see much more outreach and interest from the U.S. government, including have a seat at the table on a variety of different initiatives that the government is either in process of or undertaking relative to not only better understanding but formulating strategies and plans for the future. One important one was this RFI for near term applications that DoD put out a couple of months ago. We put in a number of different options for how quantum and D-Wave in particular might be able to deliver on that, and many of those we put in with partners, actually, so not just D-Wave alone. Then just one quick comment – it’s not just the U.S. We’re also seeing increasing interest in governments around the world, and potentially even expecting to see some large opportunities materialize first outside of the U.S.

Quinn Bolton: Excellent, thank you. Thank you Alan.

Operator: Your next question comes from David Williams with Benchmark. Your line is now open.

David Williams : Hey, good morning. Thanks for taking my questions, and Alan, thanks for all the color there. I think you answered most of my questions in that response there. But I guess one of the things that I did want to ask about is on the research that we saw that came out of China about the encryption being broken there. I know that the headline was misleading in terms of what they actually accomplished, but it sounds like it’s a bit of a double-edged sword there. It obviously shows quantum can do real things, but also coming from an adversary potentially, so can you just talk through that, what that means and how you view that overall in terms of that development? Thank you.

Alan Baratz: Yes, so first of all, I’m actually not sure today that it’s a double-edged sword in the sense that we’ve had a number of discussions with and/or inquiries from various elements of the U.S. government, and honestly, there hasn’t really been much negativism expressed around this piece of work. First of all, what I will say is that it was very interesting work. They developed an interesting approach to being able to factor large numbers on our annealing quantum computer, however still not large enough to break current cryptographic systems. The state-of-the-art for RSA today is 2,000 bit numbers, and with the technology they developed, they were able to grow it from 17 bits, 1-7, which is what we had demonstrated previously, up to a whopping 22 bits.

Okay – it’s good solid research work, it’s an interesting improvement on what had been done previously, but growth from 17 bits to 22 bits is still nowhere near 2000 bits. So while good research, good technology, interesting to note that the only quantum system today that can factor anything other than a trivial number like 15 is our annealing quantum computer, but till we’re not at the point where we have to worry about breaking RSA or [indiscernible] cryptography that’s in use today. The fact that it was done by researchers at a university in China, I would much rather the work had been done by researchers in the U.S. or maybe Europe, or other western countries, but it was research, and in the past we’ve tried not to restrict research on the system.

That having been said, I will say we do not sell access to our systems in China, we do not sell access to our systems to Chinese companies because of the U.S. government, Canadian government sensitivities there; but because we do allow for self-sign up on our Leap quantum cloud service and because of the fact that VPNs can kind of spoof where you’re coming from, it’s possible that individuals that are in countries that we generally don’t sell to might be able to get access. That having been said, given the work that had been done here, we are in the process of taking steps to tighten the access to the quantum system for self-service sign up.

David Williams: Okay, fantastic. Thanks for all that color there. Certainly very helpful. I wanted to follow up too on Quinn’s question about the NQI. I think before, you’ve spoken to some specific language that’s included here for annealing, where you felt like there could be a real significant opportunity for you, and of course given just the magnitude of the increase, do you still feel fairly comfortable there, that you could get some decent sized wins or potential there? I guess, how do you think about that for next year, once we get that authorization, what it could mean in terms of revenue or potential contracts for you all? Thank you.

Alan Baratz: Yes, so first of all, David, there are two separate bills here. One is the National Defense Authorization Act that was signed by Biden in December of last year, so almost a year ago, and that’s where the annealing and near term application language was. That has already generated the interest from DoD in annealing and near term quantum, as I’ve already discussed. The National Quantum Initiative Act is a different act, and that’s really the premier quantum research funding mechanism for the U.S. government. That’s the bill that needed reauthorization at the end of last year and it was not reauthorized by Congress, and that’s what we’re hopeful that, with Trump back in office, that will get done given that that got started while he was in office the last time. That will open up a different pool of general R&D funding for quantum technologies.

David Williams: Thanks so much.

Operator: Your next question comes from Richard Shannon with Craig Hallum. Your line is now open.

Richard Shannon: Well, thanks guys for taking my question. I’m going to follow up on the topic of NTT DOCOMO – obviously a high profile international company here, and love to see them going to production here, even without a time frame at least that you can give to us today. But maybe if you can just talk about the experience you’ve had working with them, maybe talk about the sales cycles here and how that can help inform and improve the go-forward as you go to market with other customers, please.

Alan Baratz: Yes, so NTT DOCOMO was actually a very interesting situation in the sense that they required very little help from us. They basically began by accessing the system through self sign-up. They did some early investigation, then they purchased some small amounts of time to complete the development of the application. They did some benchmarking, we provided some help along the way, and then they moved it into trial. In some sense, this is the ideal scenario for us, where we didn’t actually have to put much effort at all from a professional services perspective into the effort, and they had a great experience. They got excellent results, and they decided to move it into production. Not all of our customers have that level of technical knowledge and ability to drive the process themselves, and that’s why we still view professional services as the on-ramp to quantum compute as a service.

I could talk about how we’re seeing those time frames materialize, but with respect to NTT DOCOMO, I think it was maybe six, eight, nine months from start to where we are now.

Richard Shannon: Okay, great. That’s helpful perspective there. Alan, maybe just following up on the topic of government spending here and a couple questions before here, one of your responses was there might be earlier outcomes from countries outside the United States. Maybe you can give us a sense of where you expect that to be, and to what degree this is more research-level type of revenues or funding versus something that would be more production worthy.

Alan Baratz: Yes, so my comments about likely to see more sooner from governments outside of the U.S. refers primarily to Europe, so that’s where we are seeing the most activity right now. It’s larger numbers related to research, smaller numbers related to near term applications, so I think there’s still a view that quantum isn’t quite yet ready for prime time. That is kind of a misunderstanding that we’ve had to work hard to overcome, and still need to work hard to overcome it, and as we start to be able to talk about some government deployments of near term applications, I think it will become clearer and, as a result, easier. But the short answer to your question is Europe is where we’re seeing the most activity and larger numbers in the area of research.

Richard Shannon: Okay, that is helpful. Thanks, that’s all from me, Alan.

Operator: Your next question comes from Craig Ellis with B. Riley Securities. Your line is now open.

Craig Ellis: Yes, thanks for taking the question, and Alan and John, congratulations on the Advantage 2 progress and all that’s happening with expanding customer counts. Alan, I wanted to start on the first point there. You noted Advantage 2’s many benefits versus current systems in the favorable, fulsome consultancy commentary. The question is what’s that doing for inbound customer interest from those that want to engage with that system, and when would we expect that system to really commercialize and impact things like bookings and revenues that we’re all monitoring?

Alan Baratz: Yes, so first of all, I need to take you back six months or so to when we talked about the earlier 1,200 qubit version of Advantage 2, and what we said at that point in time was that we had just put a paper on the archive related to quantum supremacy, and that that work was actually done on the 1,200 qubit Advantage 2 system, that we had tried to get the result on the Advantage system, on the 5,000 qubit Advantage system, but we weren’t able to get the result that we ultimately got on that system. We needed the 1,200 qubit Advantage 2 system because of the increased coherence primarily, but also the increased energy scale and the increased connectivity. I think I made a comment at that point in time that arguably, even at 1,200 qubits, that system is more powerful than our 5,000 qubit Advantage system.

What we’ve been able to do now is calibrate a 4,400 qubit version of that system, showing that we’re preserving the same benefits – increased coherence, increased energy scale, increased connectivity – but in a processor that’s three to four times as large as what we got the supremacy result on, which means even more powerful. Now, the reason why I said all of this is because it’s the combination of going from 1,200 to 4,400 and the fact that we’ve been able to talk to customers about the supremacy work that is enabled by this class of processor that has generated the excitement, and a lot of that excitement right now, frankly, is in the area of materials, because that’s the basis for the supremacy work – materials simulation and research.

From an optimization perspective, while the Advantage 2 system also provides significant benefits, we really needed to get to the 4,400 qubit version to be able to get to the size problems that we’re solving in the commercial optimization arena, so we’re just now at the point where we’ll be able to start seeing some benefit from that, but keep in mind that system is not yet in Leap. It’s not yet in the cloud service. We’ve benchmarked it–we’ve calibrated it, we’ve benchmarked it, we now know that it has the same strong performance characteristics as the smaller 1,200 qubit version, so now we’ll expect to see that being able to deliver value in the optimization arena.

Craig Ellis: And it would go into Leap in the coming quarter or two, Alan, or when does it go into Leap?

Alan Baratz: We haven’t said when it will actually go into Leap, but sooner rather than later.

Craig Ellis: Okay. Second question, and thanks for all that color, on the go-to-market up-leveling, can you expand further on the productivity that you’re seeing with initial adds and some of the geos where that’s taken place, and any color on the geos where future adds will be made and how does that impact the way you think about bookings momentum and customer counts, which has noted were quite favorable in the quarter, as you look out over 2025? Thank you.

Alan Baratz: Yes, so first of all, it’s early days. As John commented, it was 11 adds but over a quarter, so these people need to come up to speed to become productive, and we still have more to do to finish building out the go-to-market team, that we expect to be completed by middle of Q1, so we’re at the point now where the go-to-market team has grown substantially, but many of the people are new and going through the learning curve. We haven’t yet seen the benefit of those people becoming productive, and we would hope that we would be seeing that starting next quarter in Q1. We are putting people, not just in North America but also in Europe, because as I commented a few minutes ago relative to government, we’re actually seeing more and nearer term interest in Europe than we are in the U.S., and commercially that’s actually true as well.

A lot of the opportunity commercially is also starting–or nearer term in Europe, so that’s kind of why we’re building out there as well. But early days, we need to get these people fully productive, and we would hope to start seeing the results as we move into Q1.

Craig Ellis: Very helpful, thanks Alan.

Operator: Your next question comes from Suji Desilva with Roth Capital. Your line is now open.

Suji Desilva: Hi Alan, hi John. Congrats on the progress here. Alan, you talked about the AI service instance, gen-AI system benefits, QPU samples. Can you elaborate on how that works, and are you seeing perhaps–you know, a lot of hyperscalers are putting in AI instances with Nvidia, AMD, are you seeing any bundling cross-selling happening or is that happening by the customer?

Alan Baratz: Yes, so there are two areas related to gen-AI model training where we have work going on today. One is there’s an approach to model training that is based on what’s known as restricted Boltzmann machines, and this approach is actually an excellent approach, but computationally generates an even heavier load than the way models are being trained today, and so it typically hasn’t been used as extensively. However, our systems, much like their native optimization engines, are in some sense native restricted Boltzmann machines, and so we’ve got a number of customers today that are looking at can we do better model training by going back to restricted Boltzmann machines and using the annealing technology to lower the computational cost and the energy costs associated with doing that, but early stages of that.

Then there are the more modern architectural approaches to model training, what are called transformer models or diffusion models, and there’s some work that’s been done actually by Nvidia in a paper they published around how to improve model training using transformer and diffusion architectures by essentially inserting in the middle of the training process sampling from a smaller, discrete latent space. Without getting into exactly what that means, smaller is ideal for us because it means we can fit it natively in the quantum computer – we don’t need hybrid, and discrete is important for us because we’re natively discrete, not continuous, on our variables. We believe that actually by using the quantum processor at that point in transformer and diffusion model training, we’ll be able to, on the one hand, generate much better models, but on the other hand, that’s a very costly part of the training process, and that’s where we think we’re also going to be able to get a reduction in time, cost and energy required to train; but in that arena, we’re not yet working with customers.

We’ve built the models. We are now working to benchmark how the QPU, the quantum processor does when inserted into that process in the model training.

Suji Desilva: Okay. Then in terms of the hyperscalers putting in AI instances for customers using Nvidia, AMD and so forth, are you seeing customers figuring out how to use your service with theirs together? Is there any bundling effort, or is that customer driven? Thanks.

Alan Baratz: The answer to that question is not yet. Currently, it’s all end customer-driven versus driven by the providers of the infrastructure to do the training. My view is that as we are able to definitively demonstrate the value of the quantum processor, whether it’s with restricted Boltzmann machines or transformer or diffusion model, that that will start to change, and the hyperscalers will become interested because they’ll have to, to be competitive.

Suji Desilva: Okay, great. Thanks Alan.

Operator: Your next question comes from Robert Aguanno with Piper Sandler. Your line is now open.

Robert Aguanno: Hi guys, thank you for taking the question – this is Robert on for Harsh. I wanted to touch a little bit on some of the comments on annealing you made, and particularly around the TAM opportunities there. Do you see potentially some solutions that previously could have been thought of as gate level opportunities potentially addressed by the annealing solution?

Alan Baratz: Sorry, could you repeat the first part of the question? I missed the context.

Robert Aguanno: Yes, just on the total TAM opportunity there.

Alan Baratz: Okay, so look – we’ve tried to point out a couple of times that the TAM for quantum covers a number of different problem areas, including optimization, including linear algebra for machine learning, factorization for crypto, differentiated equations for quantum chemistry, but it covers a number of different problem types and technology areas. One significant area is in fact optimization, and in fact many, if not most of the important, hard problems that businesses need to solve are optimization problems. I think Heather West in her quote kind of said the same thing when she said that optimization could be the killer app for quantum. Now, annealing is very well suited to solving optimization problems, and I think I made this comment a minute ago – in some sense, it’s a native optimization engine.

Moreover, there is a growing body of evidence that gate model systems are going to struggle to be able to deliver speed-up on optimization problems, and there is theory and math behind why that’s the case. There are a number of papers published a couple of years ago on this topic. There’s been more recent benchmarking done – if you go to the [indiscernible] where they did some benchmarking on optimization, comparing the D-Wave annealing system to IBM’s gate system and INQ’s gate system, you can look at the results there. On those problems, D-Wave was able to solve the problem at up to 320 variables with 98% accuracy in fractions of a second, whereas IBM and INQ could go up to about 10 variables with 75% to 80% accuracy, and taking anywhere from 100 to 10,000 times more time to solve the problems.

We think that in fact that optimization portion of the TAM is really a portion of the TAM that will require annealing quantum computing.

Robert Aguanno: Very helpful, thank you. Just a quick follow-up, more on some of the near term dynamics going on, we’ve heard that potentially some government end-of-year budget tightening could be happening, and was just wondering if that has any impact on the near term, be it the next quarter or so, just going into 2025. Thanks.

Alan Baratz: For us, no; but you should ask that question of INQ and Rigetti and other gate model providers, who up until now have been the beneficiaries of a lot of U.S. government funding – in fact, most of their revenue comes from U.S. government funding. I suspect that if there is any tightening of the budget, that will impact them far more than it will impact us because until recently, we have not–that hasn’t been our model. Our model has been commercial. We’re now starting to move into government, so that’s something we need to keep an eye on and be very careful of going forward, but currently no, it’s not an issue for us.

Robert Aguanno: Fair enough, thanks guys.

Operator: There are no further questions at this time. I will now turn the call over to Dr. Baratz for closing remarks.

Alan Baratz: Okay, again thank you all for taking the time to join us today. As I stated at the beginning of the call, interest in quantum technology is rapidly increasing, and we believe that annealing quantum computing serves as an important catalyst for commercial adoption of quantum. D-Wave is very well positioned given our continued technical advancements, highly reliable and performing quantum cloud service, progress with quantum AI solutions and remarkable work with customers on developing applications for production deployment. We continue to lead the industry in ushering in this next transformative wave of computing, and we’re looking forward to talking to you again during our next call. Thank you all.

Operator: Ladies and gentlemen, this concludes your conference call for today. We thank you for participating and ask that you please disconnect your lines.