Thank you for standing by, and welcome to the Rigetti Computing First Quarter 2025 Financial Results Conference Call. — Operator Instructions — As a reminder, today’s pro- gram is being recorded. And now I’d like to introduce your host for today’s program, Subodh Kulkarni, President and Chief Executive Officer. Please go ahead, sir.

Good afternoon, and thank you for participating in Rigetti’s earnings conference call covering the first quarter ended March 31, 2025. Joining me today is Jeff Bertelsen, our CFO, who will review our results in some detail following my overview. We will be pleased to answer your questions at the conclusion of our remarks. We would like to point out that this call and Rigetti’s first quarter ended March 31, 2025, press release contain forwardlooking statements regarding current expectations, objectives and underlying assump- tions regarding our outlook and future operating results. These forward-looking statements are subject to a number of risks and uncertainties that could cause actual results to differ materially from those described and are discussed in more detail in our Form 10-K for the year ended December 31, 2024, on Form 10-Q for the three months ended March 31, 2025, and other documents filed by the company from time to time with the Securities and Exchange Commission. These filings identify and address important risks and uncertainties that could cause actual events and results to differ materially from those contained in the forward-looking statements. We urge you to review these discussions of risk factors. 1 Today, I’m pleased to report on a number of new developments at Rigetti Computing. As previously disclosed, we have been selected to participate in DARPA’s quantum benchmarking initiative. Rigetti will advance to Stage A, a six-month performance period fo- cused on our utility-scale quantum computer concept. Our proposed concept to design and build a utility-scale quantum computer combines our proprietary multichip architecture with scalable quantum error correction or QEC codes. Our long-time partner and leader in QEC code technology, Riverlane, will collaborate with us on this project to help refine our proposed utility-scale quantum computer concept and validate the underlying technology. I’m also very excited about our recent AFOSR award to further develop our breakthrough chip fabrication technology, alternating Bias Assisted Annealing or ABAA. We will lead a $5.48 million consortium consisting of Iowa State University, the Royal Melbourne Institute of Technology, the University of Connecticut and Lawrence Livermore National Laboratory to develop a detailed understanding of how ABAA impacts the chip on a microscopic level, which aims to shed light on defects in superconducting qubits and open new avenues for understanding and mitigating them. We have also been awarded 3 Innovate UK Quantum Mission pilot awards to advance superconducting quantum computing. We will collaborate with Riverlane and the National Quantum Computing Center, or NQCC, superconducting circuits team to advance quantum error correction capabilities on superconducting quantum computers. The consor- tium will conduct ambitious QEC tests that advance state-of-the-art metrics and demonstrate real-time QEC capabilities, a requirement for universal fault-tolerant quantum com- puting. As part of the project, we will also upgrade our existing quantum computer currently deployed at the NQCC. The upgrades will include deploying a larger 36-qubit quantum 2 processing unit and integrating Rigetti’s latest generational control system, enabling improved qubit control and a fully programmable low-latency interface with Riverlane’s QEC stack. We were also awarded two additional quantum missions pilot competition projects. We will collaborate with CQC to integrate its digital chip-based technology with our 9-qubit Nuvera QPU hosted at the NQCC with the goal of identifying and understanding the key components needed for scalable QEC. We will also collaborate with Track, QRS, QControl and Oxford Ionics to create an open architecture quantum computing test bed and deliver an open specification for quantum workflows, creating a common interface between quantum software and hardware. On the technical update front, our joint paper with Harvard University, Massachusetts Institute of Technology and University of Chicago, coherent control of a superconducting qubit using light has been published in Nature Physics. For tolerant quantum computing will likely require 10,000 to 1 million physical qubits. Scaling these systems is chal- lenging because they require bulky microwave components with high thermal loads that can quickly overwhelm the cooling power of a dilution refrigerator. Optical signals have a considerably smaller footprint and negligible thermal conductivity. The team successfully demonstrated the integration of a hybrid microwave optical quantum transducer with — indiscernible — superconducting qubit. This hybrid setup enables optical control of the qubit, removing the need for coax lines and provides a promising approach to scaling to higher qubit count systems. We also recently leveraged our new quantum optimization algorithm, quantum preconditioning to address a power energy grid problem. Using a public data set representing South Carolina’s energy grid, the problem was to compute the maximum power section, a metric that informs the health and power delivery capability of the energy network. Using Rigetti’s 84-qubit Anka 3 system, quantum preconditioning was used to boost best-in- class classical optimizers. A relative advantage against the classical baseline was achieved 3 along with a high solution accuracy, highlighting the potential for quantum preconditioning to achieve quantum utility for solving tactical optimization problems. Finally, I’m pleased to report that Rigetti has closed its previously announced investment by Quantum Computer Inc. related to our strategic collaboration agreement. In connection with the closing in late April, Quanta purchased approximately $35 million worth of Rigetti common stock at roughly $11.59 per share. Thank you. Jeff will now make a few remarks regarding our recent financial performance.

Thanks, Subodh. Revenues in the first quarter of 2025 were $1.5 million compared to $3.1 million in the first quarter of 2024. Renewal of the U.S. National Quantum initiative, sales to U.S. and foreign governments and Novera are all important to future sales. Gross margins in the first quarter of 2025 came in at 30% compared to 49% in the first quarter of 2024. The lower gross margins on a year-over-year basis were due to ongoing revenues from our contract with the UK’s NQCC to deliver a 24-qubit quantum system, which has a lower gross margin profile than most of our other revenue. On the expense side, total OpEx in the first quarter of 2025 was $22.1 million compared to $18.1 million in the same period of the prior year. The increase in total OpEx was due to annual salary increases, new hires, higher FICA taxes related to the vesting of RSUs and the increase in our stock price and a reduction in the amount of engineering time used to deliver revenue. Higher stock compensation expenses for employee spot bonuses also contributed to the increase. Stock compensation expense for the first quarter of 2025 was $4.2 million compared to $3 million for the first quarter of 2024. Our operating loss for the first quarter of 2025 came in at $21.6 million compared to $16.6 million in the prior year period. We recorded $42.6 million of net income for the first quarter of 2025 compared to a net loss of $20.8 million in the prior year period. Our net income for the first quarter 4 of 2025 was driven by the noncash changes in the fair value of our derivative warrant and earn-out liabilities, which had a $62.1 million favorable impact on net income for the quarter. Derivative warrant and earn-out liabilities had a $4.2 million unfavorable impact on our net loss for the first quarter of 2024. As of April 30, 2025, following the closing of the previously announced share purchase by Quantum Computer, cash, cash equivalents and available-for-sale investments totaled $237.7 million. Thank you. We would now be happy to answer your questions.

And our first question for today comes from the line of Brian Kinstlinger from Alliance Global Partners.

In the past, you highlighted that chip tiling is expected to get Rigetti to 36 qubits at midyear and over 100 by the end of ’25 at fidelity rates of 99.75%. I’m curious, we’re just 1.5 months away from midyear. Can you speak to the progress you’re making on tiling? And do you think that these time frames are still reasonable?

Sure. Thanks, Brian, for your question. Indeed, deploying chiplets is a very critical part of our strategy because we believe that’s the real good way to scale up a number of qubits as we go forward using the monolithic die that we have used in the past, along with many of our peers. Works, obviously, and we have systems that are deployed that are doing well, but we see a significant challenge in squeezing more qubits on a single monolithic tie, which is why we deployed the chiplet approach. In the past, we have deployed 2 chiplets in a system. And then last year, we announced that we have done that with 2 9-qubit chiplets at a much higher fidelity level. We are pretty confident that the chiplet 5 approach with the 4 9-qubit chiplets sometime in the middle of this year and taking it over 100 qubits by the end of this year is going to work. All the data looks promising. We feel optimistic that we should be able to disclose our results as time goes on. I don’t have anything specific to report. I mean, as we have said in the past, our attempts with 2 chiplets have worked. All the fundamental data suggests that our attempts with 4 and eventually with multiple chiplets is going to work. This is not just important for Rigetti, but it’s frankly important for the whole industry because this is the only way we see how you scale up the qubit counts. Some other compa- nies in superconducting have tried to do larger monolithic chips and they haven’t worked quite well. And frankly, the other modalities are significantly behind the superconducting modality in terms of qubit count. So we really don’t see them coming to hundreds and thousands of qubits. So we really believe that chiplets is the right way to scale up superconducting qubits and show a path towards the fault or quantum computing systems. I hope that answers your question, Brian.

For sure. Absolutely. I have one follow-up, and then I’ll get back in the queue. I know there were some hearings recently regarding the NQI Reauthorization Act. I’m curious what you’re hearing in regards to the new funding bill and how you think about when it might get approved or what you’re hearing?

Yes. Certainly, the NQI reauthorization Act seems to have bipartisan support in our government, and there are multiple versions of the bills that have been floated both in the House and Senate. As of today, it hasn’t passed yet. And even after it passes, there is always a time for the money to be appropriated. So we are talking at least a couple of months from now before we actually start seeing money flow out of the NQI reauthoriza6 tion bill. Many of us, we, for sure, but many quantum computing companies are eagerly

And our next question comes from the line of David Williams from the Benchmark Company.

I guess maybe first, on the DARPA contract that you won, can you give a little more information about that? I know this is Phase 1, but how do you expect that to proceed? And what should we be looking for in terms of milestones?

Sure. Thanks for the question, David. I mean, the DARPA quantum benchmarking initiative project is an extremely critical project for the country. Its goal is to build utility scale quantum computing by – in the next 7 years. Really, I mean, we have used words like fault-tolerant quantum computing in the past. And essentially, it means the same thing, something that really shows value over the cost is the way DARPA defines it. Our view is to deliver a system like that, you need tens of thousands, if not hundreds of thousands of physical qubits. You need better than 99.9% median 2-qubit gate fidelity. You need faster than 10 nanoseconds gate speed and you need real-time error correction code. So once again, more than 10,000 qubits, more than 99.9% median 2-qubit gate fidelity, faster than 10-nanosecond gate speed and real-time error correction. Once you get those kinds of numbers, we believe one will achieve DARPA’s milestone of utility-scale quantum computing or what we have commonly referred to in the past as fault tolerant quantum computing. Extremely important project for the country. We are proud to be selected as the group for – selected for Phase A. 7 In a way, it’s a vetting process. Our understanding is more than 100 companies applied for that. DARPA has chosen 15 companies within the superconducting cab, along with us, it’s basically IBM and HP, I believe. And there are some other modalities DARPA has also chosen. This is called Phase A. Over the next 6 months, DARPA is going to announce the winners of the next phase, Phase B. We expect the number to go down significantly from 15. The actual number will depend on the progress that various companies accomplish. We don’t know exactly what other companies are doing. From our standpoint, we know exactly what our plan is, and it comes back to what we have publicly stated. We want to demonstrate chiplets. We want to demonstrate higher fidelity than what we have done already. Our current system, as you are probably aware, is at 84 qubit with 99% or 99.5% median 2 qubit gate fidelity, depending on what kind of gates you use. We want to bump that number over 100 qubit, but using chiplets and bump the fidelity up to 99.5% to 99.7% again, depending on what kind of gates you use. So we know what we have to do to get from Phase A to Phase B. We – I assume other companies will be doing their part as well. DARPA has said they will narrow the list down in Phase B and ultimately in Phase C. Eventually, in 1 year, 1.5 years, maybe 2 years from now, we expect 1 or maybe 2 companies are left in that process. Our goal clearly is to be in that final group. That’s where the real substantial award will be awarded to build a final quantum computer. So in a way, our view is the DARPA vetting process is a nice validation of our technology. We are proud to be part of the group that has been selected for 15. We certainly want to be part of the next group and the one beyond that. I hope that answers your question.

It does. Nice exciting to be a part of that initial group. Are there revenues that are associated with each one of those phases as you’re making contributions towards those specs? 8

Yes, there are some revenues, and we have disclosed that the Phase A is a $1 million award. We view it in a very strategic sense. I mean, as we have discussed before, our focus as a company is on R&D and technology milestones. Some of these government projects do come with sizable awards. In this case, Phase A is relatively small. Phase B, the number will increase significantly. And Phase C, certainly, the number will increase quite a bit. So our view right now is this is – we view this as a very strategic kind of a project, not really from a monetary standpoint.

Okay. Very good. And then just maybe from an interest standpoint from maybe customers that are looking to acquire the processor itself. I know in the past, it seemed like that pipeline or that interest level has picked up quite a bit. Any thoughts or colors on just maybe how you’re seeing customers now? And would you say that, that is – the interest is increasing? Or have you seen it leveling off? Anything in terms of just kind of trying to size that interest level that you’re seeing on the CPU side?

Sure. Thanks for the question. I mean, overall, quantum computing continues to be in an R&D mode. Our view is that we are still very much in the stage of developing quantum computers. We are still 4 to 5 years away from what we call Quantum Advantage, which is when you need at least 1,000 qubits, at least 99.8% fidelity median to fidelity and some real-time meter correction. Until then, it’s going to be primarily R&D, primarily driven by government contracts, academic researchers. So the – by definition, those kinds of sales are one-off sales. They are lumpy in nature. So I wouldn’t really interpret sales from any quantum computing company, honestly, right now to – in a serious way. This is all R&D going on right now. Really, the goal should be to get a quantum computer to Quantum Advantage, and that’s really when commercial sales and sales in general start making 9 sense. And we are talking at least 3 years from now, maybe 4 to 5 years from now. So until then, this is primarily an R&D kind of a situation. Having said that, I mean, there are government national labs and academic researchers who are interested in using quantum computers, whether they are on the cloud through AWS or Azure or actually getting onpremise quantum computers as we have done with Novera for academic researchers. Interest is definitely there. A lot of this organization depend on DOE funding, which is primarily the NQI reauthorization. So there is a lack of funding in general for academic institutions right now until the NQI passes and is appropriated. DoD funding situation is a little better, and that’s where we have seen some of the contracts like the AFOSR contract or the DARPA contract. But a lot of academic institutions and DOE labs certainly depend on the NQI reauthorization. So overall, even though the interest is there for Novera and on-premise QPUs, I would say lack of funding from the U.S. government is certainly holding that interest converting to orders. But again, I’ll highlight that our view is we are still very much in R&D. We focus on technology milestones. This one-off sales from government contracts as exciting as they could be, that’s not a real metric that we think should be watched at this point. I hope that answers your question.

And our next question comes from the line of Quinn Bolton from Needham & Company.

Maybe Subodh, just a quick clarification on your answer to the NQI reauthorization. Did you have a sense that, that could be signed within the next couple of months? I think you 10 said it would probably take a couple of months after the bill is signed before the funds would start to flow, but I wasn’t sure if you had given thoughts on when that may pass.

Well, thanks for the question, I mean it was supposed to pass almost a year ago, but still hasn’t happened yet, obviously. And – this is despite having bipartisan support. So in a way, it is frustrating to see the slow progress of the NCI authorization bill through the House and Senate right now. As I said, there have been multiple versions of the bills that have been floated between the House and Senate. There seems to be bipartisan support as recently as 4 days ago, I believe there was a hearing on the house floor. Again, we saw bipartisan support for the bill. President Trump has indicated that he supports quantum computing. So no reason to believe it won’t pass, but as of today, it hasn’t passed. Normally, for a bill like this, appropriations process does take a few weeks, 4 to 6 weeks is a normal number that gets used. So once the bill is passed, it still takes a month to 2 months for money to start flowing to the DOE labs and academic institutions. And that’s where really we get funded from. So it will still be a while before we see the benefits of NQLT authorization materially translate into our financials.

Understood. My first question was regarding the new award, the U.K. Innovate award, I think, was with the NCCC where you’re going to upgrade to the 36 qubit QPU with enhanced control. Do you have a time frame on sort of how long that project runs? Is that something that you could deliver that 36 QPU this year? And I guess related question, you mentioned the gross margin for the NQCC 24-bit QPU has been below average. Would you expect the 36-bit QPU to also carry lower gross margins? Or could that be revenue that comes in at better margins than the initial 24-qubit QPU sale? 11

In general, I’ll take the second part first. In general, NQCC margins are lower than the average because it’s primarily a cost sharing model that the U.K. government has. We do that despite the low margin as reported because it’s strategically very important. Right now, the U.K. government has clearly chosen us as the technology provider side when you visit the NQCC center in Harwell, which is close to Oxford, U.K., you will see that the only working supercomputer – superconducting quantum computer is rigetes right now. And the U.K. government is putting a lot of faith in us in allowing them to develop the ecosystem using our technology. So we absolutely want to continue supporting that even if it’s a cost sharing model, which leads to a lower gross margin percent on our financials because it’s strategically very important to us to be a key player in the U.K. quantum ecosystem. Regarding the first part of your question, we have disclosed in the past that even though we brought up the 24-qubit system at the NC the actual chip we use is an 84-qubit chip in that system. So it’s only really a question of cables and wires that we have to upgrade if we really wanted to just upgrade the 24 qubit to 36 qubit. So that’s one easier path, if you will, to get the U.K. upgraded to 36 qubit. But what they really would like to do and we also would like to do is once we demonstrate the chiplet approach with 4x 9 qubit at a higher fidelity than 99.5% for the generic gate and 99.7% for FSIMtype gate, we would like to bring that latest and best technology to the U.K. So ideally, we want to demonstrate that first in California in our facilities. And once we have demonstrated there, then we would like to bring it to the U.K., which would mean roughly second half of this year is when we would upgrade them to the chiplet approach and 36 qubit. But we could certainly upgrade them to 36 qubits sooner if they really want it right now using the existing UCO 3 chip, which is an 84-qubit chip. 12 So I hope that answers your question.

Got it. So it sounds like you have 2 paths. You could either upgrade the existing QPU since its 84 qubits or you could bring a tile 36 qubit once you’ve demonstrated that in the second half of the year. I think that’s pretty clear. The other question I had and maybe this is just a longer-term technical question to get to utility scale computing, you gave us sort of 4 key criteria. One of them was increasing the gate speeds. And I think today, your gate speeds are in the tens of nanoseconds range, and you want to get that down to 10 nanoseconds. What – how do you do that? Is that kind of smaller qubits, smaller circuitry? Is it just optimizing the control signals? Like how do you improve the gate speed on the QPUs?

Great question. I mean let’s talk a little bit about gate speeds first. So yes, you are right. Our Anka 3 system that is deployed on AWS and Azure right now, which is the 84-qubit system has a gate speed of about 70 nanoseconds right now. In general, if you look at all the superconducting gate-based modality quantum computing companies, our gate speeds are in the 50 to 100 nanosecond range. The numbers vary a little bit depending on the exact technology, but they are in that range. There are many knobs that can improve gate speeds. And we – before I move – get into what we are going to do next, let’s just talk a little bit about the importance of gate speed. I mean, because that often gets forgotten. I mean when we compare superconducting modality to trapped ion or pure AO modalities, – the gate speeds for those modalities are in the hundreds of microseconds. So they are like 1,000 to 10,000x lower than the superconducting gate-based modality. I’ll repeat that again. They are 1,000 to 10,000x lower than superconducting gate-based modality. And in real life, that matters. I mean 13 I have never seen any customer who has said they want their computer to be 10,000x lower. So it absolutely is critical that [ trarappon ] and Pure atom modalities increase their gate speed so to be competitive with superconducting. And honestly, I see a tremendous amount of scientific challenges ahead of them. Now coming back to superconducting side, the 1 to improve our gate speed, which was over 100 nanosecond was to move from fixed coupler technology to tunable coupler technology. We did that with Anka 2 and then we improved that further with Anka 3. That’s what got us in the 50 to 100 nanosecond range. And we noticed that some of our peer companies in superconducting gate-based modality have done similar work. After that, what we are looking at is the type of that we use has a big impact on gate speed, whether it’s Control Z, whether it’s the swap, whether it’s the SSIM, whether it’s something else. The type of gates matter because that’s what really determines how fast you can do it. But there are other important variables to how good the coupling is between the qubits and that gets into the geometry and the other design parts of the chip. So there are many other knobs out there that help us with gate speed. Our goal certainly is to continue to improve it because that’s such a critical metric that we have to monitor. I hope that answers your question.

And our next question comes from the line of Krish Sankar from TD Cowen.

This is Steven calling on behalf of Krish. The first one is for Subodh. I had a couple of questions on the ABAA development consortium. So that, first of all, that $5.48 million, I guess, grants or award for yourself and your partners, can you talk about how that award amount is to be divided amongst the partners in that consortium, first of all? And then I had a follow-up on that. 14

So yes, we don’t want to get into the details of exactly how the award gets divided at that level, Chris. We will get majority of that award out of the various partners. But I mean, in the big scheme of things, it’s honestly not that important exactly what the number is. It’s such a critical strategic project for us and the industry. So we don’t want to minimize the importance of it by getting into nickels and dimes, to be honest. ABAA is a very critical technology to scale up superconducting gate-based quantum computing. The initial results that we showed last year were very promising. The DoD, specifically the Air Force of AFOSR organization wants us to accelerate commercialization of that technology because it’s so important, some very prestigious labs, as you can see, Lawrence Riverlane National Lab, AIMS lab from Iowa and a couple of other universities are part of that because they have some key faculty members who are doing excellent research in related areas. So we think it’s a very important strategic award for the whole industry and certainly for us. So I don’t want to get into the breakdown, if that’s okay with you.

Yes, that’s fair. That’s not a problem. And the other question I had related to ABAA is, I guess, from a from a technology perspective, again, I understand that it helps to – that technology progress helps to improve the quality of the qubits. But just kind of dig down a little further into the details, does ABAA help to, I guess, increase the consistency of the behavior across different gates? Or is this helping a different vector of the performance of the superconducting gates?

In general, ABAA is – we are using it right now for what we say frequency targeting. When we have – when we build qubits, there’s a range of frequencies they respond to. And we 15 obviously want to have precise control over which frequency they respond to because once we can target them at a certain frequency, it helps us define coupling and other interactions. So ABAA helps essentially our process control of the qubit manufacturing process, gives us a precise control on the frequency. And we have papers out there that you can look at where we show the — indiscernible — spread and – without ABAA and ABAA, we really get the spread reduced by – improved by almost 10x or more. So it’s a really important way to control the frequency targeting of the qubit and get better process control of the qubit.

Got it. And just also for my follow-up question, I had one for Jeff on OpEx. Jeff, just wondering, after Q1’s $22 million in OpEx, I understand there’s a number of, I guess, seasonal increases and bonuses taken there. Just wondering for the June quarter and onwards, is like a high teens million OpEx run rate still the right number to be thinking about? Or is this low 20s level the more appropriate number we should be modeling for?

Yes. I mean I think certainly, this quarter, as I had mentioned, there were a couple of factors that caused OpEx to be a little bit high in Q1. FICA tax was one. There were some spot bonuses that impacted stock compensation expense. So going forward, hopefully, we’ll see a little bit of a reduction.

And our next question comes from the line of Richard Shannon from Craig-Hallum.

So maybe I’ll ask a question on one of the things you mentioned as well as in the press release here about optical signaling. I guess maybe if you could talk about the importance of that. And then also, I want to get a sense of when you think this enters the road map? 16 Are we getting close to that time frame? Or is this still kind of early-stage R&D?

Nice question, Richard. Thank you. Indeed, optical signaling is a very important element of our road map and for that matter, the whole industry’s road map. And the main reason for that is right now, we all use coax cables in the dilution refrigerator. That’s what sends the signals in, that’s what gets the signals out. And obviously, it works at 100-qubit level, and we think it will even work up to several hundred qubits. But then beyond that, you literally start running out of space in a given dilution refrigerator system. We are experimenting with flex cables along with many other players in the industry. And we certainly think flex cables have a role to play. So a few hundred qubits to maybe 1,000 or a few thousand qubits, we think they will switch from coax cables to flex cables. Beyond that, though, and we still need something beyond because even flex cables are not good enough to take us to the tens of thousands of qubits that we need to deliver this fault-tolerant quantum computing system or the Utility-Scale Quantum Computing system that DARPA is asking for. And that’s really where we have started experimenting with optical cables. Converting microwave signal, which is what we use to control our qubits to optical signals has been done in literature before. There have been many papers in the past. Where we came in along with our partners like QphoX, and more recently, Harvard University and MIT is showing that you can do it and actually a fiber optic cable, not just open air optics, if you will. And that’s extremely important because that makes it a lot more practicable in real life. So now we are envisioning literally a fiber optic cable going in a dilution refrigerator and the fiber optic cable coming out of the dilution refrigerator. And the physical size requirement for fiber optic is significantly smaller than even a flex cable. And that is a tremendous advantage, not only from a physical space standpoint, but also from the thermal load, 17 which converts into the cool down times and so on. So overall, we would love to switch over from coax cables to flex cables to fiber optics as soon as possible. From a practical standpoint, we think for the – right now, it’s coax for the next year, two years, maybe even three years, we will be in the flex cable regime and somewhere in the three to four years before we start getting into the thousands and tens of thousands of physical qubits is when fiber optics will start entering our road map, and for that matter, the industry’s road map. So we are definitely a pioneer in this space. We are leading a lot of the work. But once we prove it, I’m sure the rest of the industry will be looking at something like this as well. I hope that answers your question.

That answered completely, Subodh. My other question is, I’m not sure if this is for Subodh or Jeff here, but when we look collectively at the NQCC revenue opportunity as well as the Air Force, forget what the acronym is there, I think the $1 million there, what do we think about in terms of the time frame over which that will be recognized? And maybe even if you could elaborate how much we might see recognized this year?

Okay. Yes. So the NQCC opportunity, let me take that one first. That runs really for a year. So I would expect that revenue to be essentially booked over time over the course of the next year. In terms of the Air Force opportunity contract, I mean, I think when we look at that contract, we’re looking at roughly on the magnitude of – over the course of the upcoming year, it’s roughly going to probably be about $1 million. It’s a three-year contract in total. But in terms of the next year, it’s probably about $1 million.

And our next question comes from the line of Craig Ellis from B. Riley Securities. 18

Apologies if already asked. I hopped in a little late. Subodh, I wanted to start off just by going back to the DARPA Stage A and congratulations on making it to that level. But the question is this, what specifically are they looking for in Stage A? What boxes does Rigetti need to check to move on to the next phase of that program after Stage A?

So overall, Craig, the goal for the DARPA project is utility scale quantum computing in the next 6 to 7 years. They have 3 stages, Stage A, Stage B, Stage C. We have been selected for Stage A. We believe to get to Stage B, we have to execute the road map that we have publicly disclosed. So that would mean demonstrating the chiplet approach, the first 4x9 qubit at a very high fidelity like 99.5% for generic gates and 99.7% for FSIMtype gates. Assuming we do that by the middle of this year, as we have publicly disclosed is our plan and assuming we are on track to demonstrate more than 100 qubits using the chiplet approach at similar fidelity levels, we believe that should be enough to get us into Phase B because clearly, that’s a very leadership situation in the superconducting camp to be demonstrating chiplet at more than 100 qubit at 99.5%, — indiscernible — We really don’t know what other companies’ road maps exactly are and what they are committing to DARPA. So we cannot comment on that. But as far as we are concerned, we think as long as we execute the road map that we have publicly disclosed, that should enable us to participate in Phase I.

Excellent. And then the follow-up question is related to the Quantum relationship. So a quarter ago, we announced the relationship that was very significant. This quarter, we announced that the investment of $35 million has closed, so that’s further tightened and solidified the relationship. The question is this, as investors and analysts look at what 19 the next steps are for that relationship? Can you help us with some things that would be reasonable milestones from here through the end of this year and then some things that would be on the 2026 road map to help us understand where that relationship can go for the company?

Thanks for the question. Quanta is indeed a very strategic partner for us going forward. So as we disclosed, we did get the $35 million for roughly 3 million shares that they got at $11.59. But more importantly, if you look at our announcement when we made it a couple of months ago, we said that they have committed to investing $250 million over the next 5 years in the non-QPU hardware portion of the stack. So non-GPU hardware portion of the stack includes things like control systems, the dilution refrigerator, the cables, the chassis and many other components. And that’s really what Quanta is good at. I mean if you look at what role do they play in the CPU, GPU server ecosystem, that’s really what they do, and they are obviously a global leader in that space. So there, they’re bringing their expertise and their high-volume, low-cost manufacturing expertise, obviously, in addition to that. And they are going to be essentially investing on in the non-QPU hardware portion of the stack for our technology. Basically, what that means is we have to invest less in that part of the stack, which normally we would have it to do on our own. So it certainly helps us not invest R&D dollar in that. So that helps enable our focus in the QPU portion of the stack, which is basically the chip design, chip fabrication and the immediate hardware that goes around it. So long term, we can see how it’s going to play where right now, we are working very closely with them. We are essentially teaching them how to become our contract manufacturer or ODM partner, if you will, and how to build our control system that works with our stack. By this time next year, I’m pretty confident that they will be doing bulk of the control 20 systems R&D, and they would have started picking up other hardware portions of the stack along with it, reducing the pressure of us to do R&D in those areas. So I hope that answers your question of how we envision the partnership going forward.

It does. And it would seem to mean that either a R&D at current levels would refocus to something that’s much tighter and more core to scaling up Qubic quality and Qubic count or potentially R&D would go down, but it gives you a lot of degrees of freedom with how you allocate R&D dollars given the narrow focus that you’ll be able to have at that point. That’s correct. I mean it’s primarily to help accelerate our time line to market. Quanta obviously, is a terrific partner, ODM company they will certainly help us on the high-volume manufacturing side as the business really becomes commercial type business in the next 3 to 4 years.

This does conclude the question-and-answer session of today’s program. I’d like to hand the program back to Subodh for any further remarks.

Thank you all for your interest and questions. We look forward to updating you with our progress at the end of next quarter. Thank you.

Thank you, ladies and gentlemen, for your participation in today’s conference. This does conclude the program. You may now disconnect. Good day. Copyright © 2025, S&P Global Market Intelligence. All rights reserved 21