Good afternoon and welcome to Inflection's second quarter 2026 earnings conference call. Thank you for joining us today. My name is Marcus Kupferschmidt, Head of Investor Relations and Strategic Finance. Before we begin, I would like to remind you that this conference call will include forward-looking statements, including statements regarding our future financial condition, business strategy, and plans and objectives of management for future operations, and all other statements that are not statements of historical fact. These statements are subject to risks, Uncertainties and assumptions that could cause actual results to differ materially. These factors are detailed in the earnings press release we issued today and other filings with the SEC, including our annual report on Form 10-K for the year ended December 31st, 2025, which are available on our website at ir.inflection.com. We undertake no obligation to revise or update any forward-looking statements except as required by law. During today's call, we will also reference certain non-GAAP financial measures. We use these measures because we believe they provide additional insight into the underlying operating performance of the business. This non-GAAP financial information should not be considered in isolation or as a substitute for GAAP results. Reconciliations between GAAP and non-GAAP measures can be found in today's press release and in our SEC filings. Joining me today are Matt Kinsella, our Chief Executive Officer, Pranav Gokhale, our Chief Technology Officer and General Manager of Quantum Computing, and Ilan Hart, our Chief Financial Officer. Following our prepared remarks, we will open the call for questions. A replay of this call will be posted on our investor relations website along with our earnings press release and presentation materials. With that, I'll turn the call over to Matt.

Thank you, Marcus, and good afternoon, everyone. Whether you've been with us since we became public or you're dialing in for the first time, welcome, and thank you for your interest in inflection. Last quarter, I summarized Q1 in one word, momentum. This quarter, the word is acceleration. And by that, I mean the market is moving faster and the opportunities in front of us are becoming more concrete. So today, I'll start with a quick reminder of who we are. Then I'll summarize the quarter and explain what has changed in the market and why I believe inflection is well positioned. Pranav will cover our progress across computing and sensing. I'll return to discuss how we are growing the company and our outlook for the balance of the year. And Ilan will then review the financial results in more detail. And then we'll move on to Q&A. So just a quick refresher on Inflection. We are a neutral atom quantum company building quantum computers, sensing systems, and software all on one platform. Nature gives us our qubits in the form of atoms. and our job is to turn those atoms into working systems by controlling them with light and integrating the hardware and the software around them. Because our products share the same core capabilities, progress in one area strengthens all of the others. We're the only company in the quantum industry combining this breadth of products on a single platform, giving us multiple paths to commercial value. All neutral atoms, all the time. All right, so now that we aren't strangers anymore, let's move on to the quarter. Our second quarter results and the progress announced since quarter end show how much we have accomplished. We delivered record revenue of $12.6 million, up 116% year-on-year, and are raising our full year 2026 revenue outlook to approximately $43 million. Our visibility is also improving as customers commit to larger systems Thank you. Thank you. They began putting dates and dollars behind quantum. And as you will see, inflection was at the center of all of this. In May, the Department of Commerce signed a letter of intent providing for up to $100 million in proposed funding for inflection. In June, President Trump signed a quantum executive order that assigned federal agencies specific actions and deadlines across computing, sensing, and the domestic quantum supply chain. Quickly following the executive order in July, the Department of Energy selected inflection for three Genesis mission projects across both computing and sensing. Those projects bring our computing and sensing capabilities together with government and commercial partners. That same shift is underway internationally. The United Kingdom is back in quantum with up to two billion pounds in government investment. We have operated there since 2014, and recently delivered the country's first and only operational 100 physical qubit quantum computer to the National Quantum Computing Center, building on our decade of quantum work in the UK. Australia has also named quantum technology one of six defense innovation priorities for the next decade, with an emphasis on rapid field testing and deployment. We believe our growing team in Australia gives us a strong position as those opportunities develop. Governments are recognizing that quantum can deliver a level of precision and insight that classical technologies cannot, and they are moving with urgency to put that advantage to work. The commerce selection and the executive order are worth a closer look. Let me start with commerce. The headline is up to $100 million in proposed funding. That is obviously significant, but just as important is how we got there. Commerce, with deep involvement from NIST, which is the National Institute of Standards and Technology, basically really smart scientists who work for the government, took a hard look at our technology and roadmap. Following that review, Inflection was selected as one of seven quantum computing companies for up to $100 million in proposed funding to advanced commercialization. The selection is a strong external validation of our computing technology, our team, and the direction that we are taking the company. And there's something else I think is important about commerce. Historically, much of the government's work in quantum has understandably focused on advancing the science. Commerce brings a different lens. It is looking at which technologies have a credible path to commercialization and can strengthen the US technology leadership. In many ways, it was the US government determining which quantum companies they believe will scale commercially and want to deepen their relationship with. And that's exactly the criteria that we want inflection to be judged on. The executive orders reinforce the same shift. Federal support for quantum is not new. What is new is the push to put it to work. The administration's priority is very clear. Accelerate the use and commercialization of quantum. It sets real timelines for computing and sensing and brings industry directly into the work. That lines up exactly with where we've been investing and building for years. I was truly honored to be invited to the Oval Office when President Trump signed the quantum executive orders. It was a powerful moment. The message was unmistakable. The United States intends to lead in quantum. And I told President Trump that we would do everything we could to help the United States win the quantum race. We've spent years preparing for this shift. Now our job is to lead it. Inflection's advantage comes down to three things. Our neutral atom technology, the complete systems and software that we build around it, and our experience deploying quantum systems. And let me take those in order. So first, neutral atoms. Neutral atoms begin with an inherent advantage. The qubits are made by nature, and therefore they are fundamentally identical. Inflection builds on that advantage with an optical system that traps, moves, and controls atoms using light. That allows us to create large atom arrays, change their geometry for different calculations, and connect qubits without locking the machine into one permanent layout. A common optical system can control many atoms. Our systems already operate on power comparable to a few household hair dryers, and that's pretty different from building a machine that needs its own power plant. ultimately we expect a commercially useful system to fit within a few server racks and the technologies around those atoms just keep getting better advances in photonics and packaging will make the overall system more capable and economical as we scale and that gives us a more efficient path to commercial scale with a best-in-class operating profile across footprint power and cost which is a big reason neutral atoms are gaining momentum across the industry and we are increasingly seeing evidence that they will be the winning modality in quantum computing. Second, technology alone is not a product. Inflection has built the complete system required to turn neutral atoms into computing and sensing products from the hardware and the controls all the way through the software and we sell those products today. Our software connects quantum hardware with the computing environments customers already use and our teams have delivered quantum computers in the United Kingdom and Japan. and third, we know how to build and operate neutral atom systems in demanding environments. Our technology has operated on the International Space Station since 2018. We demonstrated an optical atomic clock on the Royal Navy's Excalibur submarine. We flew quantum navigation technology with BAE Systems and Kinetic, two major defense and aerospace technology companies, and we demonstrated precision timing across live fiber between Indiana and Illinois in partnership with Cephron. a global aerospace and defense technology company, and our first commercial partner for Ticker. That experience gives us engineering and operating knowledge that you only get by putting systems in the field. And that's a real advantage when customers are deciding who to work with. They want partners that can actually deliver. We have the technology, and the system's knowledge and experience are ingrained across inflection. And that is why inflection will lead in neutral atoms. Pranav will now walk you through the progress that we made across computing and sensing during the quarter.

Thanks, Matt. Our progress is racing ahead faster than ever before with the extraordinary leverage unlocked by AI tooling. In that spirit, last month, we held a company-wide hackathon supported by OpenAI who equipped us with office hours and plenty of swag. With the help of GPT 5.6 Sol and Codex, our team's executed project In quantum computing, we're translating these advances into rapid progress toward utility-scale systems that outperform classical systems for commercially relevant applications. We remain on track to demonstrate 30 logical qubits on our scale quantum computer this year, and we're happy to update today that our logical qubit demonstrations will serve three end customers, one in finance, One in energy and one in precision medicine. This past quarter, we established explicit logical qubit circuits for these customer workloads. We found new AI-discovered techniques for converting our logical circuits into physical circuits. And we experimentally demonstrated the key underlying hardware operations that power our target circuits. To make the final leap to 30 logical qubits, we are tracking two key characteristics. First, Can our system retain high fidelity even when qubits are being dynamically reconfigured? And second, can software co-design enable us to run advanced circuits with fewer than 10 physical qubits per logical qubit, rather than the hundreds or thousands that were demanded in previous proposals? We are tracking progress closely across both of these characteristics. This quarter, we also announced a new experimental record for two-qubit gate fidelity with dual species demonstrated on our scale quantum computer in Madison. Shortly thereafter, our chief scientist of quantum information, Professor Mark Saffman, co-authored a paper with University of Wisconsin colleagues proposing a novel technique for two qubit gates that paves the path to exceeding 99.9% fidelities. In parallel to our technical progress, we are preparing to offer broader customer access to our QPUs in response to the growing customer demand to run live circuits. We will communicate updates as soon as possible. In Colorado, we have advanced plans to integrate our scale QPU with AI and classical HPC coprocessing alongside the partnership we announced in November with Oak Ridge National Labs Quantum Science Center. This is a step forward to enabling a hybrid classical quantum platform which is the way we expect most customers to ultimately use our systems. The Illinois Quantum and Microelectronics IQMP Park is another exciting step. We announced last month that we are deploying a neutral atom quantum computer integrated with NVIDIA NVQ Link at the IQMP. Delivery is planned for 2027 with a new architecture designed to scale through modular upgrades to more than 50 logical qubits on the path to the system's We intend to make the system available over the cloud to a wide range of end customers. That brings me to what customers are doing with our technology. Useful quantum applications will not suddenly appear when the hardware arrives. They will be co-developed alongside it. Starting that work now allows customers to help shape the applications and the systems they will ultimately run on. To that end, on Monday, we announced a customer relationship with Eaton, a global power management leader that NVIDIA recently named among the industry leaders reshaping next-generation power infrastructure for AI. Through private cloud access to our scale QPU, our SuperStack compiler, and our quantum software capabilities, Eaton is exploring how quantum computing can be applied to difficult energy problems alongside its work in classical AI. In addition to our announcement this week with Eaton, Inflection announced last month that we have been selected by the Department of Energy for three Genesis mission projects spanning AI, nuclear applications, and quantum sensing. We are also developing energy applications through our Chicago Quantum Innovation Center. We will have more to share this fall. At IEEE Quantum Week and Quantum World Congress in September, we will share more technical progress and partnership announcements, including an update on our 30 logical qubit development. We also plan to host a webinar this fall prior to the mid-November supercomputing conference with a deeper dive into our technology and roadmap. Customers are not waiting. They're starting to build with us now. Let me turn to the sensing market. Quantum sensing addresses large markets from space and navigation to spectrum and resource discovery. and Inflection has built the broadest quantum sensing portfolio in the industry to address these markets. This quarter, we made great progress with NASA's Jet Propulsion Laboratory, or JPL, on our quantum gravity gradiometer Pathfinder program, which is moving toward deployment of the first ever quantum gravity sensor in orbit. Our teams recently visited the Einstein Elevator in Germany as we prepare for prototype testing there. The facility recreates microgravity conditions, allowing us to test the system as we continue on the path toward flight. At the same time, we're seeing growing interest in using this technology here on Earth and have started engaging with potential customers. Quantum gravity sensing can reveal features below the surface that conventional imaging cannot, including critical mineral deposits. In July, Representative Jeff Hurd introduced the bipartisan Quantum-Enhanced Critical Minerals Mapping Act, and Matt testified before the House Natural Resources Committee about how mobile quantum gravity systems could help identify and characterize those deposits at greater depth and resolution. Better information before drilling can reduce uncertainty and help the United States develop strategically important resources more efficiently. This quarter, we also extended our first mover advantage in quantum space. We announced America's Quantum Space Initiative to accelerate the development and deployment of quantum-enabled capabilities for future space systems. Many commercial companies and research institutions have reached out to join, and we will host our first meeting later this year at CU Boulder. This quarter, we introduced quantum spectrum as the first fundamental shift in RF sensing architecture in decades. Quantum spectrum demand is growing in defense applications, and we are also starting to see commercial interest. It can detect signals across a very wide range of frequencies in a compact system, giving customers new capabilities for understanding what is happening across the spectrum. Our software helps customers make sense of that information. Our quantum-inspired rapid context contract with the US Navy uses our contextual machine learning AI technology with current GPU hardware to identify what matters in large dynamic RF data sets. We are also advancing that same AI technology with additional customers, including for the US Army in GPS denied environments and for the European Space Agency across multiple data stream types. Finally, precision timing is moving into a new phase of commercialization. Ticker is selling today. Saffron is reaching global customers through its sales channel and we have secured a design win for satellite applications. We also plan to launch our next generation ticker before the end of the year, expanding the performance and applications we can address. Quantum sensing is already becoming a commercial market and inflection is built to lead it. And with that, let me turn the call back to Matt.

Thanks, Pranav. We have made strong progress across both computing and sensing and we are growing the company to keep pace with the market and with how quickly customers are choosing inflection. Since the beginning of the year, our headcount has increased by just over 25% with most of that growth in technical roles. And we don't measure success by headcount growth alone. We measure it by whether we continue to raise the caliber of the team as we grow. And we believe we've done that. We haven't only added headcount, but we've increased our talent density. And a great example of this is Dr. Joe Buck. Our new Senior Vice President of Quantum Computing Systems. Joe is a pioneer in neutral atoms, starting with his doctoral research at Caltech. He has since spent more than 23 years developing quantum and photonic systems, including, most recently, as a senior fellow at Lockheed Martin's Advanced Technology Center. Joe brings extensive experience leading large, technically complex engineering programs, as well as an entrepreneurial perspective from building and commercializing an optical startup that he co-founded. He understands the physics and what it takes to turn advanced technology into reliable systems, and Joe knows what it takes to win in a fast-paced environment. We are also expanding our presence in Chicago, Oxford, and Louisville, opening new quantum innovation facilities with significantly expanded research, development, and manufacturing capacity. We have an amazing opportunity in front of us, and we are building the team and facilities to execute against it. Turning to our financials, we delivered record revenue of $12.6 million, up 116% year-over-year. That growth was entirely organic and entirely from quantum, led by strong execution on QGG. Our performance through the first half of the year, together with the visibility that we have today, gives us the confidence to raise our full-year revenue outlook to approximately $43 million. We also remain confident that we will reach our target of 30 logical qubits this year. Ilan will now take you through the quarter and our outlook in more detail. Ilan, over to you.

Thank you, Matt, and good afternoon, everyone. I will walk you through the remaining highlights of our gap and on-gap results for Q2 2026 compared with Q2 2025. As Matt noted, we deliver record revenue of $12.6 million in Q2. Approximately 92% of Q2 revenue came from the United States, up roughly 10% point from the prior year period. As we have noted before, geographic mix will move with the timing and location of program award. Our second half outlook assumes a more normalized mix. Our gap loss from operation was $30.6 million compared with $10.1 million in Q2 2025. The increase primarily reflects higher operating expenses as we invest in our strategy along with higher stock-based compensation. Our non-GAAP operating loss was $17 million compared with $7.3 million last year. Non-GAAP operating expenses increased by roughly $12 billion year-over-year as we invest in headcount and technology development while absorbing the incremental cost of operating as a public company. Cash generated from operation was $13.2 million compared with cash used in operation of $2.8 million in the prior year period. Q2 included a temporary working capital benefit of approximately $27 million from payroll taxes collected but not yet remitted on stock options exercised by current and former employees. Excluding that benefit, operating cash burn was approximately $14 million. We expect to remit the $27 million in Q3, which will be reflected in our Q3 cash flow. Capital expenditure was $1.4 million. We continue to expense R&D as incurred and do not capitalize development costs. There is no change to our 2026 capital expenditure outlook for a measured step-up as we complete the planned build-out of our expanded facilities. We are also maintaining our 2026 outlook for a modest uptick in cash burn. which remain low relative to peers, who are deploying capital strategically to support our long-term objectives. During Q2, we completed a strategic equity investment of approximately $3 million in Oratomic, a privately-held neutral-atom quantum computing company and a customer of our quantum cores business. The investment is intended to support potential technical and commercial collaborations. who ended the quarter with approximately $582 million in cash, cash equivalent and available for self-security with no debt. Inflection is approximately 225 million shares outstanding at quarter end with substantially all shares freely tradable following the satisfaction of the lock-up release condition in April. A brief update on the Department of Commerce transaction announced during Q2. We expect to complete the definitive agreement process later this year. Under the current structure, we expect that the up to $100 million will be recorded as other income that offset eligible operating expenses. Final treatment will depend on the definitive agreement and the applicable accounting review. Back to you, Matt.

Thanks, Ilan. I'd like to briefly discuss how we are thinking about the financial model at a high level beyond 2026. The recent acceleration in market activity has improved our visibility into 2027 revenue. We expect to give more details of that when we report Q3. That increased visibility gives us confidence to continue investing ahead of 2027 and we are planning for another measured increase in annual cash burn. We believe the opportunity is large and we are investing to win. Our long-term view has not changed as our computing systems begin solving commercially useful Problems for Customers, and our Sensing Business Scales. We expect revenue to grow faster than operating costs, creating meaningful operating leverage. Let me close where I began. The theme of Q2 was acceleration. The quantum market accelerated and so did inflection. And to quote Maverick and Goose, at inflection we feel the need, the need for speed. Government priorities turned into funded programs and concrete timelines. customers committed to systems and began building applications with us. Our priorities are clear, reach 30 logical qubits this year, grow our computing and sensing business, and build the capacity to deliver at greater scale. The opportunity is becoming more concrete and so are the commitments behind it. Our job now is to execute.

Marcus, back to you.

Before we open up the call for questions, a quick note on where you can find us over the next several months. We will be participating in a number of investor conferences, including Rosenblatt, Jefferies, Citi, and Piper Sandler, as well as major industry events, including IEEE Quantum Week, the Airspace and Cyber Conference, Quantum World Congress, and Supercomputing 2026. The full schedule is shown here and also available on our website. and on a lighter note, we will soon be launching an inflection online store with company merchandise. In the interest of transparency, we'll be sure to let you know whether hoodies outperform the rest of the apparel category. As you may be aware, Matt and Pranav are hosting a livestream AMA on Reddit the week of the 24th. Several questions have already been submitted, so I'm going to start out now with one of those questions to kick off Q&A As a note, we will also continue taking questions on the subreddit and many other subreddits over the coming days. Pranav, so first question. There's been a lot of talk about the Department of Commerce selection. Can Inflection update us on what the process was actually like?

Yeah, sure. Well, thank you, Internet and Reddit, for that question. It has been a really intense but gratifying experience Department of Commerce, which, as Matt referenced, has some of the world's top physicists, engineers at NIST, went through a very intense session over the last previous few months to understand inflection technology roadmap with quantum computing and ask a lot of very proven questions. I probably lost a few years of my life in late nights and weekends, et cetera, but it was very much worth it because I think What we came out with was an outcome we're thrilled about, which is an LOI for $100 million to accelerate neutral atom quantum computing for the U.S. and ensure ascendancy alongside our allies. That took a lot of work, and again, it was alongside NIST's top physicists and beyond. It had some personal implications and reward for me, too, in the sense that I got my start in quantum at NIST. I had the privilege of growing up near the NIST lab in Maryland. And so after many years of being away from Maryland, I got to engage with a lot of the teams that I used to work with who had, as expected, some really amazing questions that helped us solidify our own paths and understand risks and what we need to do to de-risk them. And I'm feeling very confident with what we've come out of that process with. This $100 million LOI would position us to win in quantum computing period and ensure that the manufacturing capabilities for that future production remain here in the U.S. and our allies. So grateful to NIST and Commerce on that process.

Great. Thanks, Pranav.

Second question. Can Pranav help us understand the strengths and weaknesses between each modality? How can neutral atoms overcome the slower processing speeds compared to superconducting systems? What matters most?

Sure. I'll tap again into my personal background, which is when I was at NIST, I started with an internship in trapped ion systems. In my PhD, I worked on pulse-level control of superconducting systems, and so I'm familiar with the advantages and disadvantages of both those systems. What I really liked about neutral atom was first and foremost, the scalability. The number of qubits you can have in a system and inflection continues to hold a commercial record for having a 1,600 atom array. And so that is just off the charts relative to other modalities. It is true that neutral atoms and trapped ions are slower on a per gate basis. Thank you for having me. And so when you put those things together for time to solution to solving a customer application, even though the gates are slower, the number of gates that you need is much less. And on balance, those things effectively net out. So we feel very confident on neutral items being a very practical, if not the most practical pathway to getting commercial outcomes in quantum computing. And again, I also center this just in raw Do we have enough qubits to get to the crown jewels of quantum computing for commercial applications? I think neutral atom looks extremely attractive in that dimension. And that's why I kind of made the switch from trapped ion world from superconducting world to neutral atoms. Now we continue to develop software under the DARPA HARC program that crosses modalities. And so we don't want to have so much hubris to believe that other modalities will never succeed. We think there's a world where even in that scenario where other modalities have success, we continue to build an astounding business off of other modalities too. But in the near term, I'm all in on neutral atoms and everyone on our team is as well. And many of our team members, like me, switched from other modalities to neutral atoms for the scalability.

Thanks, Pranav. Operator, can you open it up, please, the line for Q&A?

Of course. We'll now be conducting a question and answer session. If you would like to ask a question, please press star 1 on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star 2 if you would like to remove your question from the queue. For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star key. Our first question comes from the line of Jesse Sobelson. with BTIG. Please proceed.

Hi, everyone. Jess Silveson with the U.S. Bancorp BTIG. Thanks for taking our questions this evening. Firstly, on the increase to Hey, guys. On the increase to approximately $43 million in revenue in annual guidance, how much of that reflects the NASA-driven Q2 strength versus in the second half improved outlook? And within that outlook, could you discuss the contribution of visibility from TICR, including that Saffron satellite design Wendy mentioned, and any expected timing from initial statements there?

Sure. I'll take a crack on that. And then, Ilan, why don't you fill in? So the increase in guidance is really a function of three things. Number one, strength in Q2. Number two, increased visibility in the form of bookings. And that is really kind of across the board. So yes, our QGG contract is a driver of that, but there are a number of other things that are embedded in that as well. As it relates to any specific specifics around ticker or saffron. I'll probably not comment on that level of granularity, Jesse, but I'll just say that the strength that we're seeing is broad-based. And as I mentioned at the end of the call, a lot of that is not showing up in numbers yet. And I think we'll have more to talk about as well in the future.

Alon, anything you'd add to that?

No, I think you summarized very well. All right.

Okay, great. And then just to kind of shift gears a little bit here on the technical roadmap, there's a lot of great information today. You identified two key steps for the 30 logical qubit demonstration, which is maintaining high fidelity during dynamic reconfiguration and running advanced circuits with fewer than 10 physical qubits per logical qubit, if I understood that correctly. What quantitative thresholds could define the success on those measures? Does achieving them largely de-risk the move from not just 12 to 30 logical qubits later this year, but potentially de-risk the move from 30 to 50 on the Illinois architecture in 2027?

Yes.

If you take a crack at that, then I'll add anything if I need to.

Absolutely. On the last piece, this is a broader step forward in de-risk to hundreds of logical qubits, and those metrics are going to be valuable for very large systems with 150 So those two characteristics, the specific quantitative metric to track on the second one is called code rate, and that just means the ratio of how many physical qubits do you need per logical qubit. What we're targeting is to be under 10, and that just means that you can squeeze a lot more mileage out of the same underlying hardware if you're under 10. The first piece, there's many ways to track this. A Thank you.

Our next question comes from the line of Tyler Anderson with Craig Hallam Capital Group. Please proceed.

Hi, gents. This is Tyler Anderson on for Richard Shannon. Thank you for taking my questions, and hello, Internet. For the modular upgrade, is this just a core swap? to increase the number of qubits? Or is there anything else surrounding that core that needs to be or could be modular as well and upgraded, not necessarily for this upgrade only, but for future upgrades with other customers?

Yeah. Hello, Tyler.

Yeah. Internet Tyler. So thanks for the question. Let me give a concrete example to Spell out what we mean. So our logical qubit results on our scale quantum computer in Colorado. We achieved 12 logical qubits in 2025, on track for 30 this year. In 2024, hit two logical qubits. That is an exemplar of this modular upgrade pathway. And specifically, in 2024, the advance that was made was the ability to reconfigure the layout of our qubits or atoms before the computation runs, but not during the computation. In 2025, we modularly added the laser systems and technologies that allowed us to rearrange those atoms during a program execution and not just before or after a program execution. So that's an exemplar of modularity. It means that there's upgrades that fundamentally advance the capabilities under the hood while retaining 99% of the same underlying base systems. and in fact, if you happen to have seen our scale QPU display that we brought to GTCDC and other GTC events alongside NVIDIA's NVQ link launch, that has our ReRanger module shown front and center. So more broadly, the notion of modular upgrades is centered on the fact that our qubits are not prefabricated. We actually always have a spare capacity of qubits. So in our 114 qubit array, Sorry, qubit system that we delivered 12 logical qubits on last year. It's actually 224 sites or atoms under the hood. And so if we want to make more mileage out of the same quantum computer, it can be as simple as taking the laser systems and swapping them out with higher power or lower noise laser systems that allow us to have access to more of the qubits that nature has given us for free. So there's certainly some upgrades that are going to be more challenging if we want to swap out the species that might be a more challenging upgrade. But by and large, the modularity comes from the fact that our qubits are free, and it's just how we control those qubits that gives rise to the types of applications we run. And we can make modular swap outs of the photonics and optics that drive those qubits.

Got it. That makes a lot of sense. And then for the benchmarking that you've open sourced, could you just talk about how that changes your go-to-market strategy or at least how it can be leveraged within one? And just any differentiation that you have in that benchmarking?

Thank you. Sure. So I think this is referring to the resource super stack library that we open sourced. And the best way to think about that is It's a vehicle for getting our customers ready to run circuits on our next generation machine. And resource estimation that we've done is extraordinarily well co-designed to our underlying hardware. There's ways to do resource estimation and compilation that are somewhat agnostic to the underlying hardware. And what we found is that for the customers that we've referenced today and will continue to reference, They want to be able to squeeze as much mileage as they can out of our next generation scale quantum computer. And so that resource estimator has some specific outlets for testing things like dual species, which is a technology that inflection leads neutral atom quantum computing on. It has things like in-place entanglement, or rather, there's technical ways of describing that too, but entangling gates that don't require motion. Those are some highlights and upsides of inflection's neutral item architecture that one can really only explore with our resource super stack, which is why we open sourced it. And then on the commercialization piece, the way we're thinking about this is a way to pour fuel on the fire of go-to-market in terms of we have so much pent-up demand, and this gives our end customers a way to start testing things while hardware access is scarcely apportioned.

Got it. Thank you. I'll hop back in the queue.

Thank you. As a reminder, it is star one to ask a question. Our next question comes from the line of Kingsley Crane with Canaccord Genuity. Please proceed.

Kingsley, your line is unmuted on my end. We may have lost him.

Go to the next question. Our next question comes to the line of John McPeak with Rosenblatt Securities. Please proceed.

All right, great. Can you hear me?

Yep. Hey, John.

Great. Hey, guys. Matt, Pranav, Marcus, and Ilan, congrats on the execution, Ray's guide. I'm a large, by the way, for a hoodie. Thank you.

All right, we'll mark you down for one.

And so you got finance, energy, precision, medicine engagements. I find that interesting. Is that generating any revenues for you guys to do those POCs or is it just kind of making them aware? It is.

All three of those.

Yeah, short answers, all three of those are revenue generating.

Go ahead, Pranav, you go first, and then we can fill in.

Sure, short answers, all three of those are revenue generating. Oh, nice.

And I know this question, just qualitatively, someone has to ask it, can you give some sense as to quantum versus sensing, either growth or just temperature-wise where we are these days?

I'll give you some high-level commentary on that, John. So we are seeing strength in both right now. And if I had to say which one I've been surprised by, I would say I've been surprised by what we're seeing under the surface on some of our quantum computing opportunities. that I think will start to be more apparent in revenue over time. And so we're seeing a lot of strength on both sides. Things are accelerating for both quantum sensing and quantum computing. But relative to maybe my expectations, I think quantum computing is coming in a little hotter than I'd expected, which is great.

Nice. It's exciting. And then just a housekeeping item on RPO. I don't know. Can we get that before the queue?

Yeah, so RPO, it will be in our 10Q when we publish it later on today. So RPO for Q2 went up to, you know, $21 million, a slight increase from our Q1 RPO. Okay. Okay, great.

And then I have one last one on the 10 to 1. physical-logical ratio that you're seeing, which is impressive for this modality. Any two-qubit logical gate error, any error metrics you can talk to there, Pranav?

Yeah, John, that's a good question. My instinct is we should save the answer to that for the upcoming events that we referenced, like the IEEE Quantum Week and Quantum World Congress, and of course, future earnings calls, just to synchronize the numbers with other numbers that we'll put out too. But I'm glad you're paying attention to it. It's a great question.

Very cool. All right. Thanks, guys.

Thanks, John.

Thank you. Our next question comes from the line of Kingsley Crane with Canaccord Genuity. Please proceed.

Hi, can you hear me?

Yep. Hey, Kingsley.

Okay, great. Yeah, glad you can. I was going to make a comment on the hoodie, too. Looking forward to get one. So either for Matt or Pranav, thinking about the space, there's both scientific progress and engineering progress needed. And so I was just hoping you could talk more about the potential for AI to accelerate innovation. You mentioned the hackathon sponsored by OpenAI, but we've also seen Pranav, why don't you answer and then I can also take a stab. Yeah, we are extremely excited about how fast things are moving thanks to AI tooling.

Indeed, the OpenAI hackathon just accelerated the pace tremendously for us in terms of adoption of Codex and 5.6 Sol. To give some specific examples, reference this in my remarks just a few minutes ago, but we have discovered a way to do operations between logical qubits more efficiently by a factor of two using an invention that actually had GPT 5.6 Sol came up with. And so a lot of the bottlenecks that used to exist on just the rate of mathematical reasoning and theory have overnight almost become no longer our bottlenecks. There's certainly an incredible role for the theorists and mathematicians at Inflection who now effectively wield armies of massive compute that turns through theorem proving and conjecture disproving, et cetera, at a faster rate. And then from a business positioning lens, We think this is a really exciting moment for companies like Inflection that hold experimental data, which I see and we see as the bottleneck to the next generation of progress. AI can prove theorems, disprove mathematical theorems, but it's going to be limited by the amount of data it has for making real-world predictions about how to build a better gate, how to get to a million logical qubits, some of the frontier capabilities that used to be extremely hard to even reason about now become something that we can pass to Codex so long as it can work with experimental data from our quantum computer and quantum sensor. So we're all in and we're positioning the work that we do to align with where AI will create value and where it's going to get pushed even forward by disproportionate advantages like experimental data. Over to Matt.

The only thing I'd add is, if I just abstracted up one layer, the new models that we have in the hands of our quantum physicists is just sort of turning them into superhumans. And so they remain the experts on how to do the quantum physics, but they now have a tool that can help them compound their efforts. And that's honestly, that's reflected across most people in the organization as we continue to adopt these tools. I continue to be surprised at how much more productive we all are becoming.

Really helpful and insightful. And just a quick follow-up. So on additional cloud access for QPUs or hosting iFQ compute on hyperscale, I'm just curious if there's any more granular detail there as that could become a more durable recurring revenue stream and assuming that's not contemplated in the guide despite some of the more recent enthusiasm.

Thanks.

Brent, why don't you go, and then I can take it practically after that. I know you're answering a lot of them first this time, but there's a lot of technical questions.

Yeah, sure. So absolutely, we're enthused about cloud access to our quantum computers, and we have been taking several steps to prepare all of our machines for Cloud Access. We think that for the next many years, there's going to be markets for both on-premise system sales as well as Cloud Access. And in that direction, there's work both in terms of setting up the cloud infrastructure, the security, which is well underway, as well as work to do new types of tooling in Superstack to account for things like user-friendly ways to do motion between atoms. One of the outcomes of our ChatGPT Hack Day over 24 hours was some of the internal winning teams assembled great ways to do compilation for atom motion and ways to visualize it, and those are going to be at the forefront of how we reach a broader set of users with tools that are unique to neutral atoms in terms of atom motion in parallel, etc., So very excited about cloud access opportunities and continuing to prepare us to open that valve up more fully. Over to Matt.

We are currently providing a lot of private cloud access to various customers and intend to open up more over the public cloud in the not-too-distant future.

Thank you so much for the time.

Thanks, Kingley.

Thank you. As a reminder, it is star one to ask a question. Our next question comes from the line of Tyler Anderson with Craig Hallam Capital Group. Please proceed.

Hi, gentlemen. Thank you for taking my follow-up. I just wanted to touch on the design win with Safran. I was just wondering what the potential timeline of when this could start ramping would be.

Thank you.

Well, I can talk at a high level about Saffron, Tyler. So, you know, what our relationship with them entails is sort of twofold. Number one is the distribution agreement where we work with them to sell tickers. And on top of that, there is a technology partnership where we're embedding their time transfer technology into ticker. and really making the two products that we each have come together as something stronger as a whole. And a good evidence of that was what we did between Illinois and Indiana showing a better than picosecond synchronization between the two sites. And so there's a lot of different opportunities with them in the hopper, and I'm really bullish on what those can bring. But I think I'll keep it at the higher level at this time in terms of what those are and when they might hit.

That's fine. And then for the revenue in the second half, is there any shape to that that we should be aware of?

Ilan, any seasonality in the second half?

I would say, as we mentioned all the time, we are expecting to have the same mix between sensing and compute. We saw the same revenue growing at the same pace. and, you know, in terms of seasonality, there is not really seasonality in the quantum. It's really depending on our contract award timing. And, yeah, sometimes you see some shifts in quarters, but we expect H2, the second half of this year, will be similar to the second half, the first half of 2026.

Okay, perfect. Thank you. Thank you.

There are no further questions in the queue. I'd like to pass it over to Marcus for any further questions.

Since we have time, I'm going to take one more from the AMA submission. Pranav, so the question was basically, can you explain in plain English why getting to 30 logical qubits matters and what this lets the computer do?

Sure. Plain English.

So essentially, we want our qubits to be as good as the bits on your computer, on your laptop, on your phone. And logical qubits are the path to making that happen. It's kind of the difference between the raw signals that your phone sends versus the actual TikTok video that your kids are watching or the Excel spreadsheet that you're downloading, hopefully Excel spreadsheet. Getting to 30 logical qubits is the stepping stone towards being in a state where our information on our quantum computer is as reliable or directionally as reliable as your laptop or as your server or as your phone. And 30, we think, is a huge step forward toward 100 logical qubits, which in turn is the point at which you start to be able to solve important problems in material science and chemistry, perhaps in AI, too. that classical computers cannot solve. And for the last several decades of quantum computing, very, very great progress was made, but it was still without logical qubits until just a couple of years ago. If we're able to hit, and we expect to this year, 30 logical qubits, I think that is a tremendous leap towards the next milestone of 50 and then 100 logical qubits. And that is the moment where quantum computing becomes everything we hope it will become.

All right.

Well, thank you, everyone, for joining us and for all the thoughtful questions from both the analysts as well as the Reddit community. We really appreciate everyone's continued interest in inflection, and we've got an important fall ahead of us, and we're going to remain focused on execution. And so, in the meantime, have a great evening, and we'll talk to you again at the various conferences we're at this quarter as well as at our next earnings call.

Thank you. This concludes today's teleconference. You may disconnect your lines at this time. Thank you everyone for your participation.