Thank you, everybody, for joining us. I'm Marcus Kuperschmidt, Head of Investor Relations and Strategic Finance here at Inflection. I want to thank you all for taking the time out of your day to join us. Today is an important milestone for a few reasons. Obviously, we are marking our first analyst day as a publicly traded company. You know, just two weeks ago, we were here at the same event location ringing the bell for our first day of trading on the New York Stock Exchange. And that was an important day for our business and our company, because that was allowing us to see the fruits of all the work that had gotten us to that point of our journey. And then we said, you know what, now we're looking forward to the next part of our journey. So what did we do at that point? It was a celebration of a successful offering. We were very fortunate to have raised $550 million in an offering equity. we were very pleased that the redemptions on that SPAC offering were quite low. And we recognized that was a vote of confidence from you, the investment community. But we understand that that trust is earned and must be maintained. So we look at today as a chance to revisit this conversation and continue to build a relationship as we work with you, our investors. Let me talk about expectations, which is obviously very important when you think about a stock. Well, we'll talk about expectations for today at this analyst event. What you should expect. You should expect to hear about leadership strength and our vision. You should expect to hear the heritage of our technology and our future roadmap. You should expect to hear about neutral atoms. You should expect to hear about our business strategy, why customers want us to be successful, and what you should not expect. You should not expect to hear about financial guidance. We are deferring that conversation to our earnings call, Q1 earnings call, which you should expect in May. So let me quickly talk about our agenda today. When I finish this and read the obligatory legal statements, Matt Kinsella, our CEO, will come up to talk about his vision. We're going to hear from our compute team talking about that side of the house. We're going to have a guest come up to talk about compute. We're going to take a quick break around 1015. We'll have 20 minute break and reconvene at 1035. At that point, we're going to talk about the sensing business and have a panel of customers talk to you. We will then bring it together for Q&A for the group and adjourn at noon. So let me before we begin, I would like to make a few legal comments. Before we begin, I would like to remind everyone that today's presentation may include forward-looking statements. These statements are based on our current expectations and assumptions and are subject to risks and uncertainties that could cause actual results to differ materially from those expressed or implied. For discussion of these risks and uncertainties, please refer to our filings with the U.S. SEC, including our Form 8K filed on February 13th, 2026, and our subsequent SEC filings. These filings, along with the replay of today's presentation and related materials, are available on the investor relations section of our website, ir.inflection.com. Inflection undertakes no obligation to update any forward-looking statements except as required by law. So with that, let me turn it over to our ceo matt kinsella all right well thank you marcus i had actually asked

marcus to turn the disclaimers into a rap or a haiku which unfortunately he did not do maybe at the q1 earnings call all right marcus already went through the agenda but just to recap you'll hear from me to start i'm really just the appetizer or the amuse bouche if you will for the main dish, which will be our computing team, Pranav and Caitlin, our sensing team, which will be presented by our CRO, Paul Lipman, and then Alon Hart, our CFO, will come up and talk to you. And then again, we've got some great panelists and guest presenters here today from SAIC, L3Harris, Dell, Safran, and then a special guest joining us via Zoom from the UK. Our guest from saffron unfortunately got caught in some travel related issues so he also will be zooming in uh his flight from rochester got canceled sadly um before i begin i have a very serious question for all of you why can't you trust an adam tyler knows they make up everything because they make up everything you're absolutely right and adams make up everything that we do here at inflection And I hope if there's anything that you take away from today, it is everything we do is based upon this core neutral atom platform. And I believe I certainly will hammer that message home multiple times. And I think that'll come across in the presentations that everybody will give today. So before I get into my introduction, I really just want to say thank you all so much for taking time out of your day to come here and learn about inflection. Our goal is to make sure this is time very well spent for all of you, and we hope that you'll learn more about the quantum market, about the neutral atom modality, and then about inflection. And most of you I have met, either in my role as CEO of inflection or in my past life at Maverick. But for those of you who I haven't met, as Marcus said, I'm Matt. I am the CEO and the founding investor of inflection. and I do want to say I am really looking forward to working with all of you over the coming years and hopefully decades. I am going to do a bit of an elongated introduction and I apologize if this is redundant for some of you who have met before, but I do think it's helpful, especially in our first analyst meeting, because it gives the history of the company in some ways as well. And so before I came to inflection, I was at a firm called Maverick for 18 years. And I wore two different distinct hats while I was at Maverick. I, for my first nine years, was on the hedge fund side of the business. So investing in publicly traded technology companies in both the semiconductor and the software worlds. And then in 2014, we decided to launch a dedicated venture capital fund. So I moved from New York to San Francisco, helped get that fund off the ground, and then was one of the partners managing what we called Maverick Ventures for the next nine years. And it was while I was at Maverick Ventures that I got really curious about quantum. This was in 2017. So I started going down the quantum rabbit hole. I met all the quantum companies that were around at the time. And the conclusion that I walked away with was that it was really too early to bet on an existing company. And maybe the smart thing to do would be to start a company from scratch. So instead of meeting companies, I shifted my focus to meeting professors at research universities and ultimately came across a gentleman named Dana Anderson. And Dana was a professor at the University of Colorado Boulder for 40 years. And when it comes to atomic physics, University of Colorado Boulder is one of, if not the best in the world. And Dana was the pioneer of a quantum modality called neutral atoms. Interestingly, in my prior trip down the quantum rabbit hole, I had not come across neutral atoms. But as I got to know Dana and learned more about this quantum modality, I became absolutely enamored with it. And that is largely because of the flexibility of the neutral atom modality. Unlike most quantum modalities, neutral atoms requires no refrigeration unit. It's all done at room temperature. And so what that opens up is a much broader possible range of applications that you can point this technology at because it is shrinkable. It is cost-downable, if that's a word. It can be field deployed. It can be hardened. And so it doesn't just need to sit in a data center and become a computer. It can be put out into the world and be a number of other things. And so the light bulb that went off in my head at that time was we could start a quantum technologies company, not just a quantum computing company, that could island hop, for lack of a better term, our way to useful quantum computing. And that is very unique to the neutral atom modality and that flexibility to be able to point it at more near term applications. So that was the initial seed thesis. I seeded the business back in 2018. I've been on the board ever since. And then I got the very unique opportunity to join full time just about two years ago. Now, all of you work in the finance industry. you all know that being at a hedge fund for 18 years is I don't know what the multiplier is but it's probably like a dog year multiplier that's like the equivalent of 100 years I thought I was going to be a lifer at Maverick I never had plans to leave and on top of that at the time two years ago I had a three-year-old and a two-month-old my wife was a native San Franciscan the grandparents lived in Marin County we had a very defined path that we were going to follow so needless to say I threw quite a curveball into our family's plans when I decided to take this role but I tell you all that. And to point out that this was not the path of least resistance, but what should have been a very difficult decision was honestly a very, very easy decision. Because this is the most exciting company I have ever seen in my nearly 20 years as an investor. And it's the most exciting industry. I really think it's going to change everything. And as I reflect back on that initial seed thesis, it's been very rewarding to see it play out honestly better than I could have hoped. We have really followed NVIDIA's monetization strategy, where we have this very core powerful neutral atom platform that we've pointed at a number of different applications and just like nvidia started by pointing their gpu platform at the gaming market and then the crypto mining market then the physics market and ultimately the crown jewel the large language model market came around we've been pointing this core neutral atom platform at timing at quantum rf sensing at the sensing market more broadly island hopping our way again for lack of a better term to our crown jewel of of gate-based fault tolerant useful quantum computing and honestly the biggest upside surprise i had a pretty good hunch we could build a good business in the sensing market when i first invested has been the unbelievable progress that neutral atoms have made in quantum computing and it turns out really until we explored this modality it hadn't been explored in in depth per novel talk more about the history of neutral atoms but they are starting to lead the pack in all of the metrics that matter and again pranav will flush that out in much greater detail so what do we do i told you about our history but what do we do again the message i want you all to take away today is we've been doing the same thing for eight years since i first seeded the business it's a very consistent a very focused and most importantly a very integrated strategy we engineer world-class neutral atom quantum computers precision sensors and software for governments, corporations, and research institutions worldwide. There's a lot to like about what we have here. We're based on Nobel Prize winning technology. And importantly, those Nobel Prizes were won in the 90s and the 2000s. This has been Dana's life work. Dana's worked on those Nobel Prize winning teams. Dana himself did not win a Nobel Prize. He is an applied physicist, so he builds the stuff. It's the theoretical physicists who get to win the prizes. But Dana's teammates won the Nobel Prizes in the early 2000s. that this technology and this company is all based on. So we are the first movers in this Nobel Prize-winning technology. We are global, and you'll see why that's important as we talk throughout the rest of the day. We've had an office in the United Kingdom since 2018. And interestingly, the UK was a first mover in quantum. It was a report that I read, put out by the UK government in 2015, that originally got me excited about quantum. And the UK has been ahead of the US in investing in quantum. The US is quickly catching up, but the UK remains a global leader in quantum. So we've had an office in the UK. we've got about 60 folks growing quickly there today in oxford we have an office in melbourne australia but the majority of our folks are here in the us our headquarters is in boulder colorado we have offices in chicago and we've deployed a system in japan and as you'll hear today i'll make this joke a couple times i like to say inflection is exo-terrestrial we've actually had our technology up in space exo-terrestrial means both in space and here on planet earth We've had our technology in space since 2018, and we'll be sending quite a bit more technology up into space in the coming months and years. We are the broadest neutral atom platform. We have a lot of Ph.D. physicists. They are the engine that keeps the wheels turning here in quantum and have a lot of the fundamental IP locked up in the neutral atom world and have serviced hundreds of quantum customers throughout our time. um i am going to spend a couple minutes here again i am going to hammer home this point that everything we do is based on this one core neutral atom technology so if you look at the far left hand side of this diagram this will be a build you will see our quantum core and that's what dana spent the last 40 years figuring out how to build it's at the core of all of our products and we are the only people who build these quantum cores to the to the uh the precision level that we can we can make them and if you see our entire our entire swath of products inside that you will find this quantum core whether it's our computers our sensing products and then we tie it all together with software and just to go one layer deeper let me just walk you through how some of these products are built how some of the underlying technology works and i think it'll come across why there is such a connectivity layer across all of them. So if we just focus on the sensing part of the product suite, at the heart of all of those products, whether it's our clocks, ticker, our quantum RF sensors, skywire, or our exact inertial sensors, you will find a quantum core that is similar to the one here on the page. And in fact, I have one in my backpack if you want to see it later in the day. And inside that quantum core reside millions of either cesium or rubidium atoms. Again, atoms make up everything, right? And in this case, what we do to build our ticker clock is we excite those rubidium atoms and we use that energy transition to create a very stable frequency reference. In fact, the most stable frequency reference that nature has to offer. The energy transition of the outer valence electron of an atom is the most stable frequency reference. And what is a clock? It's a frequency reference, right? It's gone from a pendulum to the vibration of a quartz to, in this case, the energy transition of an atom. And we use a very high-frequency laser, 778 nanometers, to excite that rubidium atom. So we have a very fast-ticking and very stable clock. And why does that matter? That matters because precision timing is at the heart of so much of the global economy. We rely on the GPS network for that precision timing for the most part. And the GPS network, as we're all reading about with the conflict in Iran and what's happening in the Strait of Hormuz, is becoming increasingly prone to being spoofed or denied. and we can provide better than GPS precision timing in a local, unspoofable manner. So we excite the atom, take advantage of the energy transition. That's how we build our clock. Very similarly, how do we build our SkyWire quantum RF antennas? Well, we excite those atoms to what's called the Rydberg state. And so you can think about that electron way out in orbit. So you now have a huge atom, and it's as excited as it can be before it becomes an ion. And when it's in that state, it becomes sensitive to the entire electromagnetic spectrum and becomes effectively an ultra-tunable, ultra-wideband antenna. Probably the biggest breakthrough in RF technologies in the last 120 years. Normally, you would need an antenna that approximates the size of the wavelength you're receiving. So for the very, very low-frequency long wavelength signals, you need massive antennas. And at the extreme, those can be a kilometer long. We can shrink that down to something the size of the sugar cube that emits nothing and can't be jammed. Absolutely game-changing technology, but very similar underlying technology to how we build our clocks. Finally, we can turn those atoms inside that core into ultra precise sensors that can sense the world around them. The way we do that is by making them ultra cold. And this is where the difference between neutral atoms and other quantum modalities really comes into play. Because it comes down to what is the definition of cold? Most people think cold means going into a freezer, and that's why most quantum modalities require refrigeration units. But in reality, cold is the lack of motion of atoms. So what we do is we hold these atoms in place with lasers such that they are moving so little, they become the coldest place in the known universe, exhibit the quantum properties, and then become sensitive to the world around them, and then we can take advantage of that. But we do it at room temperature. Again, kind of mind-boggling, but this is what Dana's been spending the last 40 years figuring out how to do. And then how do we build our quantum computers? Inside that same glass cell, we address each individual atom. Individually, each atom becomes a qubit. We entangle those atoms, we entangle the qubits, and then we perform calculations on them. But the really interesting thing is what we're doing when we build Ticker is laying the groundwork for how we create superposition in our quantum computing. In order to entangle neutral atoms, they must be in the Rydberg state. If you remember, the Rydberg state is what we use to build our RF antennas, and then the last building block of a quantum computer is it must be ultra-cold. So each of these sensing products is effectively the building blocks of our quantum computers, which is what has allowed us to accelerate our journey towards useful, gate-based, fault-tolerant quantum computing. It's the ability to engineer all these systems that are very similar at their core, and ultimately their apex is a quantum computer. And we have built and sold three quantum computers to date. And Pernod will tell you much more about that. And our software layer ties this all together. And like I like to say, we are all neutral atoms all the time, all based on the same underlying technology. I tend to like to think of things in sort of sweeping historical narratives. Maybe a show of hands. How many people here have read Technological Revolutions and Financial Capital by Carlota Perez? No one. It is a pretty obscure book. It's actually one of my favorites, though, and I could not recommend it more highly. But really, the meat of Carlota Perez's book, and again, it's called Technological Revolutions in Financial Capital, I actually couldn't recommend it more highly. It's influenced my investing decisions for a very long time and had a lot to do with why I invested in inflection eight years ago and why I came here full time. But the premise of her book is effectively that there have been five technological revolutions going back to the Industrial Revolution. They all last about 50 years, and they follow a very predictable pattern. And those five revolutions have been the Industrial Revolution in the 70s, in the 1830s, the rise of railways, in the 1870s, the age of steel and electricity, the 1920s, the automobile, and the 1970s, the information technology revolution that we're still living in today. And importantly, each of those revolutions builds upon the existing technology infrastructure, but lays down a new layer of infrastructure that effectively dictates the next 50 years. And so in the case of the Industrial Revolution, it was mechanization. In the case of the railroad, it was the revolution in transport. In terms of steel and electricity, it was the industrial scale infrastructure, mobility for automobiles, and then the digital infrastructure that we started laying in earnest in the 70s. There have been more revolutions since then. It's unclear to me as to whether AI is its own revolution or really the apex of the digital revolution. I don't think it really matters. It's kind of trying to force fit that into a paradigm. AI and quantum will work incredibly well together. If AI is its own infrastructure layer, we can call it the intelligence layer. But what I have deeply believed for many years is that quantum is this next technological revolution that is going to dictate the next 50 years of our development as humanity in close collaboration with AI. And trying to come up with a infrastructure layer that it will be laying down, I've come up with the term precision. It's not perfect, but really what does quantum bring us? quantum brings brand new computing paradigms and brand new sensing paradigms that work entirely differently than classical technologies that bring about really precision in compute and measurement at the limits of nature you're not going to get any more precise this is the smallest unit these are atoms using to compute and measure the world around us and this is truly the precision at limits of nature and make no mistake this is a global race and i mentioned we've been global since 2018. And that is very important. So it's a race against China. It's a race against a global race here. And China has been out investing the U.S. and our allies to date. And I do not believe that will persist. You've seen the U.S. list quantum, in particular, quantum battlefield information dominance as one of the six technologies the U.S. cannot and will not lose. But I did point out the U.K. as being an early adopter of quantum. If you look at the amount that the uk is investing relative to the us especially on a gdp adjusted basis per capita they're investing considerably more again i do not believe that will persist but inflection is very well positioned to capture the increase in investment from both the us and the uk and all of our allies and it's really important that the us and the and our allies win this race it's going to be absolutely game-changing to continue to hammer home this platform we platform everything we build on our core neutral atom core and and we address very large market opportunities we'll use mckinsey's number for lack of better numbers at this point in time quantum computing 130 billion dollar opportunity quantum sensing a 30 billion dollar opportunity but really as i think about this quantum computing is bringing about brand new markets that we really can't even predict quantum sensing for the most part is going to be a rip and replace cycle we're building significantly ten hundred thousand times more precise versions of existing technologies like clocks like sensors you'll hear more about that from paul later in the day and so because it's a rip and replace cycle i can imagine this market opportunity being significantly larger than than 30 billion i'll let pranav talk to this slide in more detail when he comes up here and i'll let him specifically talk about the first four rows because those are very unique to computing but let me just focus your attention on the bottom row which is enables this broad sensing market neutral atoms are unique and we are the only ones who can address this whole broad sensing market and that is because of that room temperature nature of the technology we've been at this for a very long time with a very consistent strategy and dana one you know dana worked on this technology going back to the 90s to the 2000s when i first invested in the business in 2018 it was to take that core foundational technology and begin to productize it. And we are well into the commercialization phase, which is really why I came here full-time to complete the commercialization. And we work with some of the best names in the business. Many of them are here today, which you'll hear about from during our customer panel and during the presentation for computing. But just to point out a few of these, we work deeply with NASA, with DARPA, with NVIDIA, L3 Harris, Safran, the branches of the U.S. military, the U.K. Ministry of Defense, many, many other names that we work with across the defense world, the cybersecurity world, and material science world, et cetera. And to use the term again, inflection being exterrestrial, we have done many firsts. We were the first to deploy the quantum computer at the United Kingdom's National Quantum Computing Center, which is the biggest quantum computing cluster in the world, and the only ones to complete the goals of that center on time and on budget. And we were the only foreign company selected to participate in Japan's moonshot program to build a quantum computer in the country of Japan. We were the first under the sea. The UK announced their first unmanned underwater vehicle a few months ago, and the first payload that they wanted to set sail, not just the first quantum payload, the first payload in general, were our atomic clocks and so we loaded our quantum optical clock onto the what was called Excalibur and it performed great on you know very deep deep dives. We were the first in the sky first ever to demonstrate a successful flight of a optical clock on a kinetic jet and then finally the first to put quantum into space. Our quantum systems have been operating on the International Space Station since 2018. We'll be sending more quantum tech into space in April and have announced a very large program that we're working with nasa right now to put our gravimeters into space to measure what's happening both on and below the earth's surface which paul will get to in much more detail and so just to wrap up my section before we get into the meat of the presentation i would just like to give you the takeaways that i think are worth looking out for as you hear from the rest of my colleagues and i've always had a investing framework that i've used since i started at maverick which was to look at three factors for a business, technology, execution, and financing. And I think that's not a bad way to think about what you're going to hear for the rest of the day today. So from a technology perspective, I hope this has become apparent in my opening remarks, but we have one neutral atom platform. So one stack that can spew off many different types of products in both the computing and the sensing world. And interestingly, again, that you'll hear from Pranav, neutral atoms are leading in all of the metrics that matter from a quantum computing perspective. It's been an absolutely astounding journey from kind of an unexplored dark horse candidate eight years ago to, again, leading in those metrics. And we're the only publicly traded neutral atom company, which went public, as Marcus said, two weeks ago. And we're the only publicly traded company that have demonstrated logical qubits, which are the keys to the kingdom for useful quantum computing. From an execution perspective, we are the pioneers and the first movers in neutral atoms. We have great partnerships with great customers and partners. Again, you'll hear from many of them today. And we have what I believe the right commercialization strategy, which has been our strategy since I invested. It's just executing against the initial seed thesis. And then finally, from a financing perspective, neutral atoms are inherently a capital efficient quantum modality. We have done a lot on a fraction of the capital raised and burned historically relative to other quantum companies. That said, we've raised $550 million as well in our transaction two weeks ago, which makes us incredibly well capitalized to accomplish the mission and with that i would like to tell you about i'd like to turn over to pranav to tell you about the first leg of the mission which is quantum computing so pranav over to you awesome thanks so much matt good morning i'm pranav

gokhla i'm the cto and co-founder of inflection and i'm so grateful for your time today we read many of your reports and analysis and hopefully you'll learn a few interesting things today so So by way of personal background, I was prior to Inflection, the CEO and co-founder of a company called SuperTech, which was spun out of my research and my PhD at University of And my company was deeply inspired by what the CUDA software stack did for NVIDIA's GPU And so we set forth to repeat that playbook for quantum computers or QPUs. And we got off to a fast start. In 2021, we were selling our software to the major, at that time, qubit modalities out there, superconducting qubits and trapped ion qubits. But over the course of that year, 2021, we started to learn more and more about this technology called Neutral Atoms. And spoiler alert, we got so excited about it, its scalability and its ability to address multiple markets, that by May of 2022, we decided to team up. We combined my software business with the hardware predecessor to Inflection, and that became Inflection as we know today. So I've been here now formally for four years, and prior to that, two years in the software side of Inflection, and it's been an amazing journey. And what I'm hoping to accomplish over the next 30 minutes is to tell you first what it is about quantum computing that's so exciting and makes me wake up just happy to work here every day, and what it means for the future of high-performance computing. Number two, we want to talk to you about why neutral atoms are such a promising approach to quantum computing and what they've accomplished and where we're going. And on that note, the last thing I'll leave you with is the roadmap for how we're going to execute on bringing this technology to market and how we're going to get to 100 logical qubits. That's the single metric that we're driving to and would love for you guys to be tracking. Great. So I thought it would be important to start with a big picture of what it is about quantum computing that is so profoundly different from traditional classical computing. And finding the right analogy that still is scientifically accurate tends to be an interesting balance, but I found one that I was happy with, which is that ultimately, when we look at quantum versus classical computing, you can think about rolling a dice. And I say that because at a very foundational level, the world's most challenging computational problems ultimately boil down to looking at many outcomes, like the six sides of a dice. And with traditional classical computing technology, ultimately, you can't do any better than looking at each face separately. And so with classical computing, you can try to optimize, let's say these are dollars that you'll make, and ideally you want to get $6 in this dice. But you might need several dice rolls before you get the outcome that you want. And this typifies all of classical computing at a very deep physical level, which is that whether you're looking at CPUs or GPUs, the fundamental cost scaling of classical computing is proportional to the number of outcomes. And that's a challenge because the world's biggest computational problems have not just six outcomes, but billions and trillions of outcomes. And so we need many dice rolls or many rolls of the CPU or GPU before we get the right answer, $6. And this is what is so profoundly different about quantum computing. There's two key quantum physics properties under the hood, and I'll give you your dose of morning quantum physics, which is that when we look at a quantum computer under the hood, it has two properties. The first is called superposition, which is what lets a quantum computer explore all six sides of the dice, at the same time but importantly we have to couple superposition with a property called quantum interference and that effectively lets us load the dice so that the outcomes that we want to get are heavily weighted to the point where when we roll the quantum computers dice we're going to get the right answer every time the six dollars so this of course is a small instance but But it goes to show that what quantum computing or QPUs unlock is a fundamentally different cost scaling. We have the opportunity to address dice that have many, many, many outcomes, and we can just roll the dice once instead of many, many times we need for CPUs and GPUs. And that's what's fundamentally different, this new scaling of cost that quantum computers can unlock. Now this isn't an abstract example, but the actual problems that we're thinking about have, as I mentioned, billions and trillions of outcomes sides of this dice. And importantly, we don't think that quantum computing is going to displace classical computing, but rather we think it's going to extend classical computing to unlock new markets. And there's very good precedent for this that we're seeing unfolded in front of our own eyes in the last 10, 15 years. And that's that between the 1960s and the early 2000s, CPU was the core part of the compute fabric. But, of course, as you'll appreciate better than anyone else, in the last 15 years or so, we've seen GPU stack on top of CPU and unlock entirely new applications, entirely new domains of compute throughput that were not possible previously. And yet, CPU and GPU as a combination itself have saturated on certain problems into how far they can go. And it's these problems where QPUs, or quantum computers, are the next unlock. They stack on top of the GPU and CPU stack and enable us to solve new markets, new problems that were previously intractable. And just to emphasize that this is stacking on top of GPU, Inflection has been very proud to have a number of leading demonstrations pairing our quantum capabilities with NVIDIA's GPU capabilities. And as an example, we announced just yesterday morning that Inflexion's scale-neutral Atom quantum computer is actually going to be physically on-site at GTC at NVIDIA's booth, which is very unusual, but goes to show how important it is to connect GPUs and QPUs in the same environments over a low-latency connection. And so if any of you happen to be at GTC next week, please come visit the NVIDIA booth where you'll find Inflexion's quantum computer right next to NVIDIA's latest GPU innovations. So this is the big picture, And what I want to double-click on now is just some numbers to make this more concrete. As many of you may know, one of the potential outcomes of quantum computing is that it can address cryptanalysis, RSA encryption. And to put some numbers out there, for RSA 2048 encryption, for a CPU to handle that problem, it would take 6 trillion years. Very long time. If you stack on a GPU to that, it's been found that you can solve that problem in 75 billion years, which is a heck of a lot faster than six trillion years, but still pretty darn slow. And the amazing unlock of quantum is that for these kinds of problems, we get to bring that runtime down to about a week. This is just one of many examples, but it goes to show that quantum computing is not about a 2x or 3x or even a 10x improvement. It's about fundamentally improving the way that we compute. And this can lead to many orders, magnitudes speed ups for important computational problems and i think there's no better place to look than ai to see both what breakthrough compute has brought us in the last five years but also the limits of where ai needs more breakthrough compute beyond where it's being serviced by cpu and gpu today i want to give you three buckets of where we think that quantum is going to play a key role in unlocking the next frontier of ai the first is memory and context limits as some of you may know this is one of the fundamental bottlenecks of where ai has hit a wall if you log into chat gpt or google gemini you'll find that the so-called context window how much information it can ingest is actually quite limited and it turns out that there is a quantum algorithm which we developed with our advisor Dr. Eric Anschutz called contextual machine learning that enables us to scale beyond these traditional limits of memory and context and in particular we've shown 10x memory savings versus transformers the t in chat gpt and most fascinatingly we found ways to even though this was a quantum computing algorithm built for qpus we found ways to engineer to run on GPUs, including Edge-deployed GPUs. And just to foreshadow, when my colleague, Dr. Caitlin Carnahan joins, we'll be talking a bit about some of the work that we've done on an Edge GPU platform. But the point here is that insights from quantum algorithms have taught us how to build QPUs that can overcome this limitation of AI, and importantly, deploy that, surprisingly and serendipitously, to existing GPU tech as well. This is something that we debuted at GTC last year. The second limit that I want to highlight is that models today are saturated by their performance is saturated by the quality of training data. And this is an area where we think quantum sensing is going to be an important unlock. Because our quantum sensors are delivering 100x and beyond improvements in sensitivity and precision and accuracy. And that helps AI models downstream unlock new limits that were previously held back by low-quality training data. And finally, physical AI has been incredibly powerful in the last few years. It's changed how we're thinking about robotics and autonomous. But there are fundamental limits on where physical AI has gone. It hasn't gone yet to biopharma, to materials, to chemistry. And the reason for this is that the so-called world models underpinning physical AI are not currently able to capture the underlying electron-electron interactions that are beyond anything in medicine and materials, etc. And QPUs are that unlock. They enable existing world models for physical AI to capture the full set of physics that are needed, a thousand X more performantly than GPUs and CPUs can, to enable new frontiers of physical AI. So, the way that we see this all coming out into the market is as follows. The thing I want to emphasize here is that inflection is not just building towards what happens in a few years. We deliver value to our end customers today, and that's through quantum-inspired techniques that were built for QPUs but ported over to GPUs and CPUs, and those have enabled our customers to have new types of improvements for navigation, for sensor fusion, and spectrum awareness in parallel we're also building quantum readiness services on cpu and gpu to get the end users the enterprise engaged on where quantum computing is going around year 2028 we expect the first applications of fault tolerant gate-based quantum computing to emerge in full force and we think that the first set of applications are going to live in materials and this sounds like a very scientifically abstract world, but in fact you and I interact with materials every day. Imagine if your laptop battery could last 10 years or if your car could have fuel efficiency that's 100x beyond what it is today. This is the domains in which we see quantum computing playing a role in new frontiers of materials. Shortly after quantum computing addresses important applications and materials, we see exciting applications to QPU accelerated AIML, biopharma, chemistry, drug discovery, and a variety of different optimization use cases a little further down the road. Importantly, we need the right technology substrate to get to the quantum computers that unlock these types of applications. As I mentioned, over the course of 2021, I became deeply enamored with neutral atom qubits. And I say this as a bit of personal background, as someone who started my career in a trapped ion lab at NIST in Maryland, and then in grad school worked on superconducting systems. But what I got incredibly excited about with the neutral atom technology is that the atoms are completely identical. And that means that our qubits are completely identical. There's no manufacturing defects. It's the gift of nature. It also has impressive scalability that I'll show you concretely and quantitatively soon. But we have the ability to load thousands of qubits into a single of these neutral atom cores. and those qubits have very long what are called coherence times, which is basically the lifetime of the qubit. Importantly, the way that we effectively wire our qubits is not electrical wires that have to touch every single qubit, but we wire with light, with photons, and that gives us exceptional levels of control, and it also enables any qubit to talk to any other qubit in a constant number of steps, which has an important unlock for how fast we can get to this concept of logical qubits. And inflection has a heritage, as Matt mentioned, of building this neutral atom technology for many years through Dana's leading work at this company. I wanna pause just to reflect on what makes neutral atoms and what I just described different from our peers in trapped ion and superconducting, with whom, again, we share great admiration and my own background comes from those communities. Maybe the way I can describe this is that neutral atoms and trapped ions, this next row over from neutral atoms, are actually quite similar. They're both atoms. It's just that the ions have a charge and the neutral atoms are neutral. They don't have a charge. And so in many of these rows, there are similarities. Both of our platforms have been able to demonstrate applications with logical qubits, room temperature operation, high connectivity with our qubits. But the one thing that makes neutral atoms very different from the trapped ion modality is that we can get thousands of qubits in a single core, which for us obviates quantum networking we don't need to hook up many cores to get to important quantum applications we get there without networking and that's something that is very much leading the pack in that top row on neutral atoms there are superconducting platforms that have also crossed about a thousand qubits but at the scale that we're operating at it's a neutral atom advantage and then finally i just want to echo what matt mentioned which is that we have the opportunity to address with the same technology that we're building towards quantum computing towards the markets of timekeeping, RF sensing, inertial sensing, gravimetry, and all of these platforms feed directly into our quantum computing roadmap. And that's the beauty of when we joined Inflection from the software perspective, we realized the amount of leverage in this platform is unlike any other qubit technology. So I've talked about this word called logical qubits a few times. Many of you may already be familiar with it, but I really want to double-click because, to borrow Matt's expression, logical qubits are the keys to the kingdom. And anytime you hear an algorithm requiring a certain number of qubits, what they really mean is logical qubits. And it gets weaved into a lot of advanced math and physics, but at the core, it's a concept that we're all familiar with in our daily lives, which is when we look at data centers or hard drives, it takes about 30 terabytes of storage across many different AWS centers or partitions to get to one terabyte of reliable storage and the reason for that is that there can be outages in certain regions there can be reasons to have redundancy through the RAID format if you've ever bought a hard drive before and so it's very familiar in the classical computing world that if we want one terabyte of reliable storage we're going to need more of the actual disks that store that and that's the exact same concept that underpins logical qubits there's a distinction between the physical information and the logical information and in particular for quantum computing you can conceptualize that there are many qubits or atoms under the hood let's say 30 or more but we weave them together from unreliable individual qubits into one logical qubit that is highly reliable and it's in fact the same technology that is used in classical computing with data centers, LDPC codes, error correction, etc. And it's important to emphasize this because getting to 100 of these so-called logical qubits is the unlock for the most transformative applications of quantum computers. And so if you walk away with nothing else from this segment, just know that we are driving hard towards getting to 100 logical qubits and getting not to unreliable qubits that can't be used in customer applications, but to getting to these reliable qubits that are important for material science, for AI, for chemistry, for pharma, and beyond. And just to mention now, breaking it down, what does it take to get to logical qubits? We can really break it into three pieces. This is my kitchen recipe for getting to logical qubits. The first thing that we need to get into logical qubits is we do need to have many of these underlying physical qubits. This is the quantity axis. And this is a open source data set on the right hand side, where all we've done is identify which points belong to neutral atoms and which points specifically belong to inflection. And I want to point out that this is a logarithmic scale on the y-axis. And you can mention, you can notice that one of these curves and one of these colors is far beyond any other compute trajectory. And that's the neutral atom trajectory. And in particular, inflection has set the commercial record period for the number of qubits. We've done 1600, which is many times what other qubit modalities have been able to do. We've seen progress in areas like trapped ion, but they're still stuck at about 100 qubits. This is a 16 to 1 ratio against other qubit modalities. So that's step number one. We need a lot of physical qubits, but we need those physical qubits to have high reliability or high fidelity operations between pairs of qubits. And so this is the quality axis how good are these qubits and I want to point out a few things here so first off lower is better we want to be down and to the right that's the error on the y-axis the first thing that you'll see is that neutral atoms are the new kid on the block this qubit technology began on this plot in year 2010 and that was when our chief scientist of quantum information professor Mark Safman for the first time ever ran an experiment showing that neutral atoms can be used for quantum computation with entangling gates. And since then, it's been very much a down into the right curve. And if you really look closely, you can see that there's a dramatic fall in error here, which is good. And honestly, I attribute this to Matt Kinsella for being the first venture capital investor to look at neutral atoms as a modality and decide that this technology has real room to grow. And what we're seeing today is that inflection, this neutral atom quantum computer is operating in the same fidelity regimes as superconducting qubits which have been along for a much longer time and specifically within neutral atom commercial systems inflection holds the record for this critical metric of fidelity we've achieved 99.73 percent user-facing fidelity of our operations between physical qubits so this is the quantity and quality the hardware side of the stack but we need a third equally if not more important ingredient to stitch these physical qubits together into logical qubits, and that's the software stack. How do you go from this raw physical qubits to logical qubits? The answer is this platform that we've been developing called SuperStack, and SuperStack is the CUDA for quantum. It does to QPUs what the CUDA software stack did for NVIDIA's GPU stack, and in particular, we've published results that show that using SuperStack to apply to a set of physical qubits, we can extract 10x performance improvements, which is effectively delivering the impact of years of physics progress, but virtually through the software stack. And I want to emphasize that this platform has been so successful that, in fact, we've sold it not just to neutral atoms, to ourselves, but to other competing qubit modalities like silicon spin qubits, like trapped ions, like superconducting, other entities are turning to us to help them go from physical qubits to logical qubits. So that's SuperStack in a nutshell, and my colleague Caitlin will touch on this as well. So that's a recipe for logical qubits. We need many physical qubits. We need high-fidelity gates, and we need the software stack to stitch together. I want to emphasize that we have now put these three ingredients together to deliver logical qubits. And so these are three of our key results that we put out in the last few months and years. In 2025, last year, we published this result on the left-hand side, which is that record for the highest fidelity commercial neutral atom system in terms of the entangling gates. Shortly after that, this was just published now a couple months back. This is a paper that we jointly worked on with NVIDIA as our collaborators and co-authors. And it's the first ever time that an application in material science has been run on logical qubits. And it exemplifies how QPUs and GPUs are going to be co-processing important workloads together. And the final one is our magnum opus of our recent work. This was published or released as a preprint just a few months back. And it shows 12 logical qubits taking a dramatic step forward towards 100 logical qubits. And as I'll show, this was actually delivered ahead of schedule. Under the hood, we've got 114 physical qubits on the system with all-to-all connectivity. And the key takeaway here for me is that inflection has now reached a stage where we're integrating the scientific advances in each of those three ingredients into engineered deployed systems that process as reliable quantum computers. just to double click on this magnum opus paper that i mentioned this is the one that we put out in september of 2025 and i want to quickly highlight three of the really important takeaways from this paper the first is that many of you may have heard of shore's algorithm for a while this is the quantum algorithm that could break rc encryption what we showed in this paper is for the first time ever running shore's algorithm with logical qubits and the reason that we're doing that is to send a message to our friends and collaborators in the cyber security community that the moment where quantum computers cause us to have to change our encryption standards is happening. In fact, it's already happened. And it was a message to the industry to be prepared for the impact of quantum computing. In the middle is our 12 logical qubit demonstration. And I'll emphasize here that it wasn't just about making 12 logical qubits appear and then walking way, we actually ran arithmetic on these 12 logical qubits and in a specific way that actually accelerates the timeline for material science and chemistry applications. And on the right-hand side is a novel approach that we unveiled in this paper for improving the ratio of physical qubits to logical qubits. I mentioned this 30 to 1 ratio previously, but we see room to make that even potentially 24 to 1, which is very different than other roadmaps out there that require 1,000 physical qubits per logical qubit. It's not just about having the scientific results integrated into machine, but it's also about getting these machines to market. And so Inflection is proud to have delivered and sold two quantum computers already. One is the first system installed and operational at the National Quantum Computing Center. It's a 256-atom, 256-qubit array. That's out in oxford in the uk in japan we've delivered to our customer a 500 qubit qpu to the institute for molecular sciences and last year we were very proud to announce with the illinois quantum micro electronics park a planned delivery of a 50 logical qubit system upcoming this is the on-premise side of the stack where we expect to have continued growth in on-premise sales But we do see an emerging and booming cloud market, and we've used our Superstack platform to put our scale quantum computer on the cloud. And in fact, it's also available through the NVIDIA CUDAQ cloud, with whom we've done demonstrations of our live machine to the supercomputing conference. We've delivered to a national bank that's a customer and a number of other entities that are getting really excited about quantum computing. So this is what our roadmap looks like. And I just want to emphasize the logical qubits as the key metric that really matters. When we put out this roadmap in early 2024, we declared that this would be the year that we got two logical qubits. And on track, we got to two logical qubits. We had a more detailed version of a roadmap that also had specs for 2025. And at the time, we had projected that we would get to eight logical qubits in 2025. And as I alluded to earlier, we were very proud to end up hitting 12 logical qubits in an acceleration of a roadmap, exceeding a roadmap. In 2026, we've projected that we will achieve 30 plus logical qubits. And 2028 is that commercial inflection point where we expect to achieve 100 plus logical qubits. And this keeps going. We see progress to 1,000 logical qubits and beyond. and every logical qubit that we add is going to dramatically expand further the set of applications that we can address. So that brings me to the end, and the thing that I want to pause on is there's so much going on at Inflection that is exciting about quantum computing that I didn't even have the chance to speak about a lot of the ingredients even beyond what's on the screen. So I thought I'd at least pay a little bit of lip service to some of the other key milestones that if you Google search, you'll find more about. but there's a number of different metrics that we're tracking with respect to dual species technologies, large qubit arrays, records, and measurement fidelities. And in fact, our chief scientist of information, Mark Saffman, was proud to win the Bell and Ramsey prizes, which are among the highest honors in the field of AMO physics just a few months back. In the middleware and architecture side, we've been proud to develop these new approaches for efficient logical qubits, integrate with NVQ-Link. Again, if you're at GTC, come visit the NVIDIA booth to see our machine live. Launch contextual machine learning and launch a software library in collaboration with JPMorgan Chase for more efficient approaches to error correction. And finally, I want to emphasize that everything that we're doing is we're doing it for the commercialization. And so we've been proud to announce these three systems. And as well, just on Monday, we announced that our SuperStack quantum software platform has been adopted by several of the US's leading national labs. So these are the systems ingredients towards commercial applications and that is what we're driving towards in quantum computing at Inflection and on that note I'm very excited to welcome my colleague Dr. Caitlin Carnahan to talk about quantum software and where we're going with applications of quantum computing.

Thanks Pranav. Good morning everyone. My name is Caitlin Carnahan. I'm the vice president for quantum software at inflection by way of extremely brief sort of level setting i am a computer scientist and quantum physicist by training and i'm really excited today to basically piggyback off a little bit to kind of you know take this wonderful overview that he's given of our computing efforts go up the stack and start to talk about what is software doing for us and what are the applications that we are unlocking with software. So to start with, let me go back to a point that that Pranav made earlier. So I think there's there's no doubt that AI is a revolutionary technology. And I think it's fair to say we haven't actually seen what the ultimate impact of AI is going to be. So it's really natural to wonder where is sort of the frontier for AI and how do we anticipate quantum going beyond that frontier? So to kind of start that conversation, I think it's helpful to kind of get a sense for where we expect that AI and classical computing are going to continue to fall short. So the first area here that I have listed is combinatorial explosion. So this is the domain of complex optimization. Generally speaking, when we think about optimization problems, we have a number of variables or choices that we're trying to optimize over. And as we add more variables or choices to our problem, the space of solutions that we need to explore is going to grow exponentially. So to just have a little mental model that might be helpful, we can think about the task of delivery routing. So this is a task that is mathematically structured as a Hamiltonian path and what that means is that let's say I'm a truck driver and I have to make some deliveries and I have five stops to make there are over a hundred ways that I can plan out my route if I have to make ten stops there are over 3.5 million ways I can plan out my route if I have to make 15 stops which is far fewer than the number of stops that typical delivery person makes in a day I have over a trillion ways to plan out my route so as i add more complexity to my problem as i add more choices to my problem my problem space is growing exponentially and if you think about for a classical computer to just kind of brute force check all those possible combinations and compare them quickly get to the point where that is just not possible the second domain is is intimately related to this problem but specifically i'm talking about the limits of physical simulation so similar to to the delivery routing problem, when we want to consider classical simulation of quantum systems, we also experience this exponential sort of explosion of the spaces that we are going to need to consider as we add more units to our problem, whether they're atoms, whether they're molecules, whether they're biological structures, we have that exponential growth. The third is a little bit different. It's with respect to data limitations. So if we just step back a little bit and think about what AI does. Fundamentally, what you're doing when you're training an AI model is you are asking it to essentially reverse engineer a function that is corresponding to the data that it has seen. It's trying to learn or deduce a function. If you don't have a lot of data, or if your data is not high quality, or if for whatever reason your data is not encompassing the entire space that the function needs to define, your AI model runs the risk of just not learning the correct thing. And this, of course, is potentially a, well, it is a challenge in several key critical areas, especially in defense. So for quantum solutions, what we're considering here is not just looking for patterns in data like we do with AI, but trying to exploit the underlying structure of those complex problems, trying to get to the underlying mathematical structure, the underlying physical structure, and try to exploit that structure in order to solve the problems more efficiently. So now we have a sense of where quantum can add value on top of AI. But to unlock that value, to unlock these advantages that we talk about with our quantum hardware, we need to develop a quantum native software stack. So our team consists of subject matter experts that range from computer scientists to physicists to machine learning experts, and even experts in key application areas like quantum chemistry. And we leverage this combined expertise in order to develop proprietary tools and software techniques, including our optimized compilation, which Pranav mentioned earlier. And those all go into our flagship quantum software platform, SuperStack, which also serves as the user-facing kind of external gateway to our quantum computing system scale. So when we consider classical computing, one thing that is maybe a little bit in the weeds, but I think it's important to point out, is that for classical computers, we traditionally leverage this concept of deep abstraction in order to decouple the software layers from the hardware and the middleware from the top of the stack. But the truth of the matter is that it is just too early for us to be doing that in the field of quantum computing. And in order to move towards quantum advantage, essentially what we believe the strategy is going to need to be is to be a thoughtful, integrated hardware-software co-design effort. So when we work together towards these applications, It's not just the software team that's sitting all alone in a room. We're actually talking with our hardware team about what is doable, and we integrate together. Inflection has the right team. We're the right size. We have the agility, but we also have the discipline in order to identify these shortest paths to quantum advantage and to chart out the course accordingly. And we're not doing this all in a vacuum. I tried to work in some kind of joke about our ultra-high vacuum glass cells here, but I can't make it work. But what I mean to say is that our quantum software team is working with external customers. We have customers that span from our national labs into potential industry end users. And we, as sort of evidence of the expertise that we are bringing in our software stack, SuperStack is actually the quantum software that powers many quantum platforms across multiple modalities. So these days, much of the conversation has really focused on, as Matt mentioned earlier, the sort of the crown jewel of quantum technologies. So that would be the far left end here where we're talking about large-scale fault-tolerant applications that are giving you quantum advantage in these key domains. But really, what we focus on is not only that sort of long-term goal, but also where can we realize quantum value today? So there's a couple of key milestones that I'll kind of highlight here. In the nearest term, one thing that Pranav introduced for us was this concept of quantum-inspired AI, which is something that our quantum software team is working on. And the key concept behind quantum-inspired AI is that we are essentially leveraging principles learned, lessons learned, from the study of quantum information, from the study of quantum computing, to rethink how we design AI models for classical systems. So what I mean by that is that, you know, some of the choices that we make about the actual architecture of the model that is underpinning these applications, they could be unintuitive, but we find that they can yield advantages, especially in, as Pranav pointed out, the study of, or the understanding of long-running complex correlations in data. And we find that there are a number of applications that benefit from these quantum-inspired sort of techniques to model design, especially in the spaces of sensor fusion and spectrum awareness. And I'll just point out that this is a wonderful complement to sort of match the power and precision of our quantum sensing capabilities. In the midterm, we are also focused on the development of hybrid quantum workflows. So here we're talking about application spaces and middleware that support the ability to use classical computing to do what it does best, but also to supercharge its abilities with quantum computing. But that goes both ways, actually. So not only can the quantum computer, you know, take the hardest kernel of the problem and help out the classical computer, but classical computers can also help to accelerate the maturation of quantum computers. And one way that that is possible is through, for example, the use of classical machine learning and classical computing techniques residing alongside the quantum computers to help with some of the challenges that we have in scaling up fault-tolerant quantum computers, especially with respect to routing and decoding challenges. But then, of course, our long-term goal, our ultimate goal, is the development and the demonstration of fault-tolerant applications. So these large-scale applications that are going to demonstrate scientific and commercially meaningful results in several domain spaces, so some of which we'll highlight today, materials design, drug discovery, and large optimization just being a few. So there's a lot to talk about here in terms of what value we can get, but the key takeaway, I would say, is that on our sort of path to getting to those crown jewel applications, there are a lot of stepping stones of value that we can demonstrate. At Inflection, we strongly believe that quantum hardware development must be motivated by realistic and customer calibrated use cases. We are not building a quantum computer just for its own sake. We are building it to solve the hardest challenges that we face as a society. And I think that's a really key point. These challenges are found in every sector. So, of course, we're going to be excited when we get to a point where we're solving these challenges with quantum computers. But even further, the ability to solve problems at that scale is going to inspire us to imagine new challenges that we couldn't have even considered before. So I would like to take a minute to highlight some of the work that we're doing in these spaces. And I'll also call on some of my colleagues to help me finish this list. So the first is in the health care and biosciences space. So we were excited recently to announce that we were selected to participate in the final phase of Welcome Leap's Quantum for Bio Challenge, Q for Bio. This is a supported challenge, the aim of which is to accelerate the applications of quantum computing in the health sciences space and also to figure out where is quantum going to disrupt the health care sector within the next three to five years. our project which is a collaboration with mit and the university of chicago focuses on finding biomarkers so biomarkers are clinically informative feature sets that are constructed from high dimensional multimodal data biomarkers are used in the healthcare space in order to customize treatment to do diagnosis and to essentially chart a course for the the care of a patient. However, the ideal biomarker is a small set of features. And one of the reasons why we want a small set of features is because they're more biologically actionable, clinically translatable. They can be interpreted. And so we have this big problem where we may have a lot of features that we can choose from, not a lot of patient data in order to sort of optimize the selection of those features. But we do want to make sure that we get the best set of features that we can so we can frame this as a combinatorial optimization problem essentially we're searching through a vast number of feature combinations we want to find the best ones and that space is of course going to explode exponentially and even more concerning is the fact well not concerning but it's challenging is that under underneath all of this data that we're using to analyze, you know, the potential biomarkers that we can leverage and to help to treat a patient. We have a number of different data sets, some of which are represented up here, which actually are all correlated. They're coming from the same biological mechanisms. So we have these really complex correlation structures in this data that make it an even more compelling problem for quantum computing. so in our approach what we did is we developed hybrid quantum classical algorithms um in order to essentially take some of the hardest parts of that that optimization problem send them to the quantum computer and then let the classical computer take on the rest of the problem when it's back into a solution space that is that is tractable for the classical computer we're not avoiding these complex correlations we're tackling them head on with quantum and this is already paying dividends we have unexpectedly compact interpretable feature sets that come out of the results of our study we have cross data set performance that is that is robust across multiple data sets that we try and we have a uh an outlook on what the um the pathway looks like from hardware-aware development to clinical translation. But the most exciting thing that I like about this project, I love to point this out, is that you can see up there, our partners are MIT and University of Chicago. At the University of Chicago, we are not partnering with quantum physicists. We are partnering with computational biologists. We are partnering with practicing clinical oncologists. We are partnering with the eventual end users of this technology, and they are helping us to validate it on problems that actually matter. So then I can move on to the energy sector here. So within the energy sector, again, we're talking about, as you'll notice, there's kind of a theme here in the types of problems that we tend to tackle, although they appear in many different sectors, problems that kind of show up as huge optimization problems. So within the energy sector, we are tackling the optimization of energy delivery. And I'm very happy to point out that actually this work here, which is featured in code, was ARPA-E's first ever quantum technology contract. And within this work, what we are doing is we are helping to understand how the challenge of energy generation, energy transmission, contingency planning, how you can get all those resources together and actually optimize that planning in such a way that the quantum computer can help accelerate that problem. Our work is focused on hardware-software co-design, so this is a joint effort between our software team all the way down through our hardware team, thinking about not only what can we do at the software layer to solve this problem, but how can we really direct the development of our hardware to make the most out of the algorithms at the top. Another thing that I would like to point out about this effort, again, as before, is that we have a number of partners in this initiative. So we're partnering with Argonne, the National Laboratory of the Rockies, as well as EPRI. And we have a number of stakeholders that are buying in and helping us to understand that the challenges that we're tackling here are not just challenges that our quantum physicists are just imagining in their heads. We are talking to industry end users. We are talking to partners to understand what would this actually mean for you? What are the relevant formulations of the problem that we should be tackling just to make sure that we are really finding that path to value? And then lastly, I'll just point out one of the applications that's closest to my heart. So I am a condensed matter theorist, which, if you're familiar, is basically in the space of material science. This is one of the most promising applications of quantum computing. Many of you who are close to this space may have heard the sort of the famous Richard Feynman quote that nature isn't classical. So if you want to simulate it, your simulation better be quantum mechanical. So this is really what we're trying to get to the heart of here. When we are considering the design of new materials, we have to understand and understand the behavior of interacting particles. Usually, especially for strongly correlated systems, we have particle-particle interactions, and we can even have them at higher orders, that we just can't sweep away. We can't approximate the way. We can't sweep them under the rug. We have to consider them in order to get a physically meaningful understanding of the system. And this is exactly what we are tackling with our early fault-tolerant demonstrations here. So our 2024 demonstration, what we did is we looked at the two-logical qubit single impurity AIM model. So the Anderson impurity model is a model for understanding how magnetic impurities occur in metals. And we essentially took a version of this problem, did the ground state estimation, logically encoded. And while this problem is small, what I want to really point out here, that's really the breakthrough from this effort is that we were able to do this in a logically encoded manner. And Pranav, of course, impressed upon you the importance of those logical encodings. I would also like to highlight that this work was done in collaboration with NVIDIA. Through this work, we supported integration of our SuperStack scale platforms with CudaQ. This is also another really important point to make. The availability of these GPU accelerated workflows within this work allows us to use GPU acceleration in the early parts of development. We can design our circuits and we can simulate our circuits on these ultra-high-powered GPU-supported frameworks and then take those exact same circuits that we simulated, we studied, that we prepared, and port them over through SuperStack into scale. So now I would like to ask one of our dear friends, Dr. Chris Powell, to come up and comment a little bit on national security applications.

Thank you so much, Caitlin. Dr. Powell, I'm known as Dr. Q on LinkedIn. I'm the quantum information science lead for SAIC, but also chief scientist and fellow. And if I get this correct, then we get this. And this is awesome. Why am I here? I'm here because of the mission. And why is quantum relevant to the mission? Because we can do things now with quantum that we could not do before and cannot do with classical computing, classical sensing, or other types of things. Over the last year, it's been interesting. So about a year ago, I would talk to customers all across the services and ask them, do you have any plans in quantum or are you interested in the technology? And the response you always got was, what's a quantum? And in the last year, they're now actually seeking us out and knowing that because of the improvements we've been able to make in algorithm performance and mission relevance, they're really seeking us out because we're solving problems that cannot be solved traditionally. You heard Caitlin talk about a number of problems in material science and financial aspects as well as drug discovery and things like that. They tend to experience three main problems. One of them is exponential degradation in that if I do an incrementally more amount of calculations or more data in the problem, the number of calculations goes through the roof. A number of these problems also have a lot to deal with electronic structure or problems that mimic electronic structure, and I'll talk to that in a second. And then the other one is the need to do a lot of uncertainty estimation. And when you combine those three things, your algorithms get rather ridiculous. And one of the interesting aspects to that is, you know, you've heard perhaps that the cell phone you carry in your pocket is essentially more powerful than a Cray-1 supercomputer from the late 1970s. And while the computation is different between the two, it is actually a relevant example. And what we're getting with the scale computer coming out of inflection is comparable to turning this into what's different than the Cray-1 being a regular supercomputer. And part of the aspect of that is if you look at the Frontier supercomputer at Oak Ridge, and I'm a chip designer, board designer, and I come from that HPC kind of OEM background. All the wiring inside that supercomputer is equivalent to nine and a half trips to the sun and back every nanosecond. How do I partition everything I need to do in something like that? You have to change the device, meaning the physics, and you have to change the math. And that's what quantum gives us is the ability to kind of invert that idea and get us to faster calculations and more mission relevance faster in that kind of an environment. And that's because it allows us to get beyond Moore's law in terms of transistor count, Denard's scaling, the size of a transistor and the voltage that's needed, and Amdahl's law related to the optimization that's needed for those at a task level. So by changing the math, we're putting everything into superposition and being able to take a look at the problem as if it was an electronic structure problem in that we're looking at energy levels of everything that's related to the problem. that kind of a matrix we heard earlier was called a hamiltonian and that's not the hamiltonian up the street in broadway but the one that actually in this case gives us more efficiency in solving problems um and you know we also change the fundamental of the the device that we're working on and that we're getting past the transistor and the device now is an atom and we're able to exploit that at the the level of the uh the electron so you know we're looking at one of the significant mission problems that we have is in missile defense and in related areas. In really bad day scenarios, you're going to be confronted with thousands of potential threats and decoys. And how do you deal with that? It's in the warfighting job, that's a function of what we call magazine depth. How many interceptors do you have? How much can I apply force in order to resist that and then convince, by the way, the enemy not to do it again because I'm going to come after them in that kind of a setting. It becomes very, very difficult. That problem explodes exponentially very, very quickly. In the progress we've been able to make with quantum algorithms and in quantum computing, we've been able to take three to four orders of magnitude of wall clock time out of the calculations for that and massively increase the accuracy and the decidability of the courses of action that we can propose in that environment. What does that mean? In an ICBM, if you're getting from point A to point B, you have about 30 to 45 minutes pre-launch to intercept in order to be able to do something. In today's hypersonic cruise missiles, you have 10 seconds, if you're lucky, end-to-end pre-launch to intercept. In our calculations and what we've been able to do in modeling, we can get sub-10 seconds out of that. That's why in SAIC we say act now on quantum computation and quantum sensing behind it, because we can also get much better data coming into the system, instead of invest right from that perspective. And that's why our customers are now taking a hard look at this and why we're taking a very hard look at it from an implementation standpoint. Quantum is one of the technologies I take a look at. Next-gen computing is another. Next-gen energy, because we have to deal with the fact that not everybody can have a gigawatt data center in their backyard, and also bioscience and other types of applications, which is a little less from the idea of how do you manipulate biology and more how can I exploit it, such as DNA memory, DNA computing, in order to come up with models that are better. So in the environment that we're taking a look at here also, the cryptographic existential threat is very significant. I have to be able to protect against any kind of intrusion on my data, any kind of spoofing of it, any kind of denial. I'm confronted in my warfighting environment with a tremendous amount of that kind of a threat. And I have to be able to deal with that in real time and get past it. In the mission space we're talking about, this mission space is largely unpredictable, but it is modelable. And so the quality I can give the warfighter is I can make their warfighting system more resilient. Today, I don't have that ability. Classical computing doesn't give me that. Now, the warfighter will always get the job done. But then what it is is a compromise on the courses of action they have to put together, the decision criteria they have, and how many resources they have to deploy, how much magazine depth they have, and how they have to approach a warfighting problem. They'll always get it done, but why not hand them an environment where I can get things done in seconds, where I can give them trust from an agentic AI using CML that gives us an ability to have a far more assured and trusted set of calculations for this. This is what we're being able to offer the warfighter with quantum now, and it's why it's such an advantage and such a wonderful thing to be able to work with inflection on this in this kind of an environment. We're really getting to the point now where with the quantum sensing that we've also got coming up later, we can get much better accuracy and precision of information in what we're getting into the system and what we're computing. Now, it's important to take a look at from a roadmap perspective also how things are working. Today we're working with qubits, a d equals two, so the behavior essentially of a hydrogen atom. You have with Spintronics the ability to exploit higher d, and every time you're exploiting an additional dimension in that regard, you're getting additional orders of magnitude of computational throughput. This is what the strength of it is, and this is already built into SuperStack from that standpoint. So as we're increasing the logical qubits, as we're also increasing the ability to exploit D, we're getting to the point where we can really hand the warfighter solutions to problems that were considered intractable before from that standpoint. So that's why we're here. We're here from the standpoint, certainly from the warfighter, certainly from the intelligence asset. But there's lots of comparable mission problems in this space that our customers face every day In financial transactions, this country is the world's reserve currency, and in that kind of an environment, it presents obviously a tremendous attack surface, but it also presents a very complicated trading model, and with these types of computing systems, we can get past some of the implications of that to handle more of the threats and be able to prepare more proactively for an environment and a trading environment that's far more complicated. it. In medicine, we can get to simultaneous discussions of toxicogenomics at the same time we're taking a look at drug candidates. If you go to the FDA and you're trying to certify a drug biologic or device, you get a four-section form. It's a single page, and this is all the things you need to fill out about your new drug and things like that. What's behind that are tens of thousands of pages, gigabytes of models and movies and everything, and an MD has to sit at FDA and sort through all of that as they're trying to figure things out? Why not be able to provide them an ability to simultaneously kind of sort through a tremendous amount of that? And again, that's what this gives us from the mission perspective. I will then now turn it back over to Marcus, and I appreciate the opportunity to talk to you all. You bet. Thanks so much to Dr.

Powell and to Dr. Carnahan. I encourage you all to chat with them during the upcoming break and really gave a sense of how inspiring the mission is and how critical it is for our way of life, frankly, and how important it is to the nation's future. I have one final spotlight for you, which is what we're doing in AI. And I want to spotlight three customers that Inflection has been working with over the last few months and years. The first is the U.S. Army, where we have contracted with them on a program called Sapient, which stands for Secure AI for PNT. PNT is

position navigation timing as matt alluded to there has been a massive increase in the denial

of gps in both civilian and military territories and what you're seeing here is a demonstration of our sapient quantum software platform and so if you see here there's this red circle that is tracking this asset and after gps has been denied or spoofed we basically checked this box to turn off gps in the simulation environment you see that that red circle is still tracking the car how does that happen it's because our sensor data fusion platform called sapient is integrating together inputs from computer vision from inertial sensors from altimeters and many other types of inputs and it's this technology that the army has been working with us to take our quantum inspired ai and deploy it to existing solutions and i mentioned this earlier but we've been taking this technology to the edge the insights that we took from quantum computing actually apply on edge gpu platforms where we get a better performance per watt of power consumption per megabyte of memory etc and this is that nvidia edge jetson platform in one thing i want to foreshadow perhaps for the next session is that we're also building the sapient platform to be forward compatible with quantum sensors these inertial sensors in the next few years will start to become quantum inertial sensors. They will have quantum radio frequency receivers and optical atomic clocks like Ticker or quantum clock to perform the timing in position navigation timing. So it's an important part of how our commercialization strategy has evolved, which is we're building the software stack for both classical and quantum technologies, and that positions us well to then integrate our hardware into our existing software at our customer sites. With the US Navy, we've been working on a similar edge-deployed quantum-inspired solution called Quark, quantum-inspired rapid context. And this is enabling us to do spectrum awareness, collision avoidance at the edge using GPUs. And then finally, in Europe, we're working in the UK with the European Space Agency to plot to taking these edge-deployed AI solutions based on quantum technology to space, to satellites. And in fact, again, planning to integrate insights from quantum sensors and actually hardware integrations quantum sensors. So that brings you to the close of our computing side of the house. Inflection has built strong leadership in neutral atom computing. We've set the commercial record absolutely for the number of qubits, that's 1600 physical qubits. That's the quantity axis and on the quality axis we've delivered 99.73 percent reliability, which is the neutral atom commercial record for user facing entangling fidelities. And we've put that with our proprietary software stack, SuperStack, to deliver on 12 logical qubits, which was ahead of our roadmap. We projected eight logical qubits in 2025 and beat it. This year, we are targeting 30-plus logical qubits with 100-plus logical qubits in 2028, and as we're getting out to market, we're keeping our eye on near-term, medium-term, and the longer-term in terms of how to build in revenue dollars. In the near term, we are taking our technologies to existing GPU-edge platforms with quantum-inspired solutions. In the medium term, we're integrating with GPU. And our crown jewel is this opportunity with fault-tolerant quantum computing to which we're climbing rapidly and taking massive steps every month. Well, with that, I thank you for your attention and look forward to chatting with you during the breaks. And I'll pass it to Marcus.

Thank you, Pranam. Really appreciate it. So we are a few minutes off our agenda. So what we're trying to do is take our break now and let's reconvene at 1040. Thank you.

I'm Paul Littman. I'm the chief revenue officer for the company. I've been with Inflection for five years. In fact, actually, I was looking at the calendar this morning. Today is actually my five year anniversary with the company. so very fortuitous timing. I spent a couple of decades prior to coming to inflection in the cybersecurity market. I've played a number of roles here at the company and honored now to be leading our go-to-market efforts. So I'm going to talk to you this morning about our quantum sensing business. Quantum sensing, kind of simply put, are systems that measure the world, that sense the world with greater precision and capability than classical state-of-the-art. And this is very much a dual-use business. We address both defense and commercial applications. As Matt showed you earlier, McKinsey forecasts this to be a $30 billion market by 2040. But actually, when you think about the potential for replacing and upgrading existing infrastructure, We believe the market opportunity is far greater than that. And by virtue of the fact that we use the neutral atom modality, we are able to address the key market segments of timing, of RF sensing, sensing the electromagnetic spectrum, inertial sensing, measuring motion, acceleration, rotation, and gravitational sensing. And it's the same core underlying technology that we use in our computing business, as was talked about earlier. So quantum sensors, Inflections quantum sensors, are addressing critical customer needs today. We're truly delivering quantum advantage with sensors, solving real-world problems with real-world products today. Ranging from GPS, which we all rely upon every day, which undergirds much of the world's economy. It's increasingly spoofed and denied, not just in conflict zones, but even here domestically as well. I'll talk a bit more about that in this presentation. And in conflict zones, it's not just the GPS part of the spectrum, but in fact, pretty much the entire spectrum that is increasingly congested and contested. We're seeing the emergence, as Dr. Powell talked about, of new forms of threats like hypersonics, that's the impetus behind the Golden Dome architecture. And as we see in the news literally every day, the rise of drone warfare to devastating effect in the Ukraine and now in the Middle East. And then space, which has also become a domain of conflict, is an emerging and potentially enormous source of economic growth. We're seeing the rise of mega constellations, talks about putting data centers in space. I think even Elon Musk said maybe putting a quantum computer in space. I think that's a little bit further down the line. And as we think about going back to the moon for exploration and resource discovery and mining of lunar resources, the asteroids and beyond. And the challenge that classical sensors have in addressing these problems and these opportunities is that they've effectively reached the limits of their capabilities. So this is where quantum has such a critical role to play and where inflection has such a terrific opportunity for growth and leadership. So we'll start with timing. So most of us, when we think about GPS, we think about location. You take your phone out, open Google Maps, where am I? But actually, underlying that, timing is the fundamentals of GPS. It's a timing signal that's then triangulated to determine where you are. And timing is, of course, important for PNT, Position, Navigation, and Timing. It's in the name that's used both commercially and also for national security. But in actual fact, GPS timing undergirds everything from precision agriculture to power generation and distribution, telecommunications, financial trading. We're sitting here today at the New York Stock Exchange, the world's premier stock exchange. Every trade, every transaction that comes through the global markets has to be precisely time-stamped. And that requires GPS timing. So if GPS were turned off, if GPS were fully disrupted, we wouldn't be able to trade. We wouldn't be able to communicate. We wouldn't be able to know our location. We would have incredible challenges distributing power. The economy would come to a halt. In fact, it's been estimated that if that were to happen, if GPS were completely denied, it would cost the U.S. economy over $1.6 billion on a daily basis. So the part inflection has to play in addressing this problem is our optical atomic clock product ticker. This is a commercially available product. It is a 3U rack-mountable system. It's 100 times more precise than GPS timing standards. And we've demonstrated this, as Matt mentioned, as I'll talk about in more detail, in a wide variety of operational environments, from on the land to on the sea, under the sea, in the air, and we are working towards space qualification of this technology. So fundamentally, Ticker enables our customers to communicate, to operate, to navigate, even in a scenario of complete GPS denial. We're partnering with various companies. We have folks here from some of them here today and joining us remotely. And I'll point out a recently announced partnership with Safran. Safran is a global defense and aerospace leader. And we're partnering with Safran, who has a large business in P&T and timing, to accelerate the commercialization of Ticker, to amplify our go-to-market. Saffron has a large go-to-market organization, a large customer install base, and to drive the upgrade of legacy classical technologies with Ticker and other quantum sensing modalities. So we're using these partnerships to accelerate the growth of the business on a global basis. We've had customers who've demonstrated and validated Ticker in a variety of use cases. So, as I mentioned before, as Matt talked about, we partnered with the UK's Royal Navy to demonstrate the operation of an optical atomic clock in an unmanned submersible. And the reason that this is important, if you think about long-duration operations under the sea, you don't want your submarine to have to surface to synchronize your clock to GPS to get a timing signal. And so from a national security perspective, this will enable longer duration submerged operations. We did a similar demonstration in the air in combination with BAE and Kinetic, demonstrating again for the first time the operation of an optical atomic clock in flight for similar PNT reasons. And then lastly, at the end of last year, my colleague Max, who's sitting at the back here, worked with our partners at Quantum Corridor in Illinois to demonstrate picosecond level timing synchronization between tickers over tens of kilometers of urban fiber. So this is not a lab-based experiment, but really in the wild through the Chicago and surrounding areas. So demonstrating and laying the foundation for future quantum networks and for synchronization of distributed workloads for AI and high-performance computing. So this is the Ticker roadmap. Ticker, as I say, is available today commercially, Ticker Prime, our first offering. We manufacture that in our Colorado facility and also in Oxford in the UK. It is a 3U rack-mounted system and essentially designed to be rip and replace for existing timing standards. The next release of Ticker will be Ticker C. And the fundamental focus, I should say, of our roadmap is to reduce the size, weight, power, and cost, what's referred to as swap C of Ticker. And the same will be true with our other products. So the next release of Ticker Ticker C will be a lower bomb cost version of the clock, also in the 3U rack mountable form factor with a wider temperature operating range to enable use cases in more broadly deployed environments. The subsequent version of TICKER, TICKER HD, which stands for heavy duty, and TICKER S for space, is a further reduced form factor and ruggedized and hardened for operation in military deployed environments and also for deployment in space. And then lastly, by the end of the decade, our roadmap objective, using the work we're doing in photonic integrated circuits, is to reduce TICKER down to card scale. And this will enable us to go after the miniature atomic clock and the compact atomic clock markets, but with a product that is 1,000 times more accurate than current standards. So fundamentally, again, the roadmap, reducing size, weight, power, and cost, unlocking new applications and new growth market segments. So I'll turn now to talk about radio frequency sensing. The chart you see here on the right is from a recent NATO report and illustrates the complexity of the electromagnetic environment. Every asset you see here on the screen is emitting RF radiation or receiving RF radiation or attempting to intercept or disrupt RF radiation. And the challenge that we have with classical technologies is, firstly, that classical antennas scale with the wavelength of the signal that you're trying to communicate with. So typically, say, for long-range communications or for communicating with submarines, we use very long wavelength, low-frequency signals. And those antennas can be of order of a meter, tens of meters, in some cases hundreds of meters in length. Antennas, classical antennas, are inherently narrow bands, so it's not unusual, for example, to see a military vehicle covered in antennas for addressing different parts of the spectrum, and they're inherently jammable and detectable. And if they can be detected, they can be destroyed. In fact, we've seen in Ukraine the average lifetime of a monostatic radar has been reduced to just a matter of minutes. Once something is detectable, it can be targeted and eliminated. So the net result of this for our warfighters is increasingly challenging for them to communicate, to retain situational awareness, and ultimately to remain safe. Inflection is addressing this by pioneering in the breakthrough field of quantum radio frequency sensing, also referred to as Rydberg sensing. And as Matt said, it's been described as the biggest breakthrough in RF technology for over a century. And essentially, the innovation here is we are replacing antennas with atoms. We tune these atoms with lasers into the Rydberg state that we talked about before, the same Rydberg state approach that we use in our quantum computer. And these atoms can be tuned across the entirety of the electromagnetic spectrum. And the atoms are in a small vapor cell, literally the size of a sugar cube. So we're taking an antenna that could be the size of a football field and replacing it with something that would fit in the palm of my hand. And because these are just atoms in a vapor cell, it's electrically silent, so it can't be detected. And because it's exquisitely tunable, it's inherently resistant to jamming. So this opens up a broad array of national security use cases. We're working with a variety of national security customers in the US, UK, and Australia, some of whom you see here on this slide. So, for example, we could take an array you see in the graphic of our Skywire system here, a 4x4 array, and we could tune each of these sensor heads to a different part of the spectrum for broadband sensing capability. And fundamentally, everything emits, so we can sense broadband spectral coverage in a very covert way to detect signals of interest. we could tune each of those sensor heads to the same wavelength for very precise geolocation of emitting signals and many other applications beyond fundamentally the same approach from a roadmap perspective as i talked about for ticker reduction of size weight power and cost and adding additional capabilities and performance benefits over time and this is really an area that inflection is leading the market with really tremendous innovation. You'll hear more about that from our sensing panel in a few minutes. So I'm going to turn now to talk about inertial and gravitational sensing, which are essentially two sides of the same coin. In inertial sensing, and in PNT more generally, the objective is to develop a fully self-contained autonomous system that can enable low-drift, long-duration navigation in any domain, land, sea, air, or space, and to do so entirely independently of external signals or maps. So if you're relying on an external signal, well, that signal could be spoofed, or it could be jammed, or it could be otherwise interfered with. And if you're relying on a map, well, maybe you're operating in a sensitive area for which maps are just not available or not reliable. I've talked about the T part of PNT, where inflection has innovated with ticker. We've also demonstrated a number of other world firsts. We created the first ever ultra-cold matter, Bose-Einstein condensate in flight. That's an important component of inertial and gravitational sensing. And we also demonstrated the world's first continuous cold beam inertial sensor at sea. and we'll be doing further demonstrations of this technology with the UK's Royal Navy later this year. As I say, same underlying technology we use for inertial sensing enables gravitational sensing. Some of the most important changes on the earth or under the earth are invisible to cameras. So think about water table levels, for example, or critical resources under the earth, or gravitational dynamics that are important for national security or navigation purposes. And the reason for doing this in space is it enables broad coverage. You can see, obviously, a lot of the Earth's surface from orbit, but also enables continuous coverage as you orbit the Earth to continue to take measurements and understand changes over time. Inflection has a very deep history and legacy and experience in space. We developed the core physics systems, in fact, a number of physics systems for NASA's Cold Atom Laboratory that's been operating on the International Space Station since 2018. And as a result of this work, we were selected by NASA JPL to develop the core physics systems for the Quantum Gravity Gradiometer Pathfinder mission. And that's a multi-year program. We've booked $20 million of business to date on this program. And the ultimate goal of QGG is to put an exquisitely sensitive gravitational sensor in orbit that goes far beyond the capabilities of classical systems. But as we think about space, it's not just about gravitational sensing. You could think, for example, about the benefits of putting a quantum radiofrequency sensor in orbit. And I've already talked about the work we're doing towards space quantifying ticker. So space represents an exciting growth market in resource discovery, in national security, in infrastructure. And again, as we think about space exploration and ultimately mining the moon and the asteroids and beyond, this is a real opportunity for growth and leadership for inflection. On the topic of space, I'm delighted to say that Inflection is one of just a handful of companies selected for the Golden Dome MDA shield, IDIQ, and we envisage a range of applications for Inflection's technology across the Golden Dome architecture, from QRF for detecting the signature of hypersonic vehicles in flight. As Dr. Powell said, you just have seconds to react, so being able to identify these systems early is critically important. Ticker for enhancing radar capabilities and ensuring other systems can operate even in complete GPS denial. As Caitlin talked about, with contextual machine learning, the ability to understand anomalies and recognize patterns at the edge in real time. and with quantum computing, the ability to optimize for asset deployment, utilization, and ultimately for decision-making. So to wrap up, Inflection's quantum sensors address critical needs for customers today and established a foundation for large-scale infrastructure markets tomorrow. Our leadership in core neutral atom technologies enables us to deliver quantum advantage in the key markets of timing, RF sensing, inertial and gravitational sensing. And it's the same core technology that we utilize in our quantum computing business. We can ingest data from our quantum sensors into edge deployed contextual machine learning for advantage today. And then over time, as we bring quantum sensors and quantum computers together into integrated solutions for even greater benefit tomorrow. It's a dual use market. we're addressing the defense and commercial sectors. We've demonstrated use cases with customers in defense, commercial, and space sectors. We're partnering with companies like Safran to accelerate the path to commercialization, to go-to-market scale, and ultimately to upgrade existing deployed bases of classical technologies. And our roadmap fundamentally focused on reducing size, weight, power, and cost to unlock scalable markets and scalable applications. I do have just an example here of what I'm talking about. So this is a piece of integrated technology developed by inflection. This is a prototype for a future inertial sensor. So taking something that would have been a few years ago, a lab bench scale, to something that would have been a dorm room refrigerator, to ultimately integrating all of the components into something that fits in the palm of one's hand. And this is something that is made in our Colorado facility. So thank you very much for your attention. We're going to turn now to our customer panel. So I'd like to invite Chris and Tom up to join us here. And then we have a couple of folks joining us remotely. Tanner Cheek from Safran and Sir Grant Shapps, former UK Secretary of State for Defense. And I'll let the AV guys hopefully will have a zoom here momentarily. Here we go. Hi, Tanner. Good for you to thank you for joining us.

Hi. Good morning. I apologize. My flight got delayed this morning because of maintenance. But thank you for the flexibility and thank you for the invitation. Happy to happy to be here.

Thank you for joining us. And we'll we'll figure out how to get Grant on here as well momentarily. Maybe I can tell some physics jokes to fill the time.

Where's Matt when you need him? Exactly.

The temptation is so great, but I'm going to resist. Yeah, maybe we'll do that. So I'd like to, first of all, start with Chris. Chris, I'll let you introduce yourself.

Yeah, so I have been running the – been head of quantum sensing and quantum computing and quantum networking at L3Harris for the last eight years and been working with a lot of research customers throughout the country and throughout the world with these types of applications. And Tom.

Yeah. Hi, Tom Treacle. I work at Dell Technologies in the Federal Services Department. And we actually, my background is test and evaluation, but I also am heavily involved in taking technologies to market in the mission space. I do a lot of support with the Department of War, et cetera. But my real interest is being here to support what I really feel personally is a step function, really change in the way we do business for solving mission problems at the Tactic Ledge.

Great. And Tanner, I'll let you two go next.

Hi, good morning. I'm Tanner Cheek. I'm the vice president of sales and marketing for ST4D, which is the Saffron timing business line. So I lead our time reference, time distribution business line for our global and U.S. commercial customers.

And Grant, glad we were able to get you in. Let you introduce yourself. No sound. I don't know if Grant can see us, but Marcus, maybe if you could message him. We can see you, Grant, but we can't hear you. Well, while we're waiting for the audio to get figured out, I'll just say that Sir Grant Shamps is a former UK Secretary of State for Defense and now co-founder of an innovative defense tech company, Cambridge Aerospace. So maybe we'll start with an opening question. So quantum sensing, as we've talked about, is moving from the lab into real mission deployments. What do you see that's changed in the last few years that makes this kind of an important time for quantum sensing to find its way into the real world? And Chris, maybe I'll start with you.

Yeah, I didn't know that Matt was going to bring this up, you know, the book that he mentioned, but I'm going to use a similar analogy. Henry Ford once stated that if he'd asked people what they wanted, they all would have said a faster horse. So when we look at these types of technologies for quantum and emerging technologies like this, the fact is that when the classical sensors are pushed to the physical limits that they're capable of achieving, you've got to start looking at it from a different perspective or a different angle. And so the market has kind of moved from being, you know, just physics, you know, related. We've been able to prove the physics. And, you know, at this point, we've now been able to push this capability into achieve things that we couldn't achieve otherwise. And so, like, the White House recently has come out with quantum as being one of their top, you know, solutions. and now we're seeing the deployment of these types of sensors in various applications where we can provide value that we couldn't otherwise.

Great. Tom?

Yeah, so from my perspective, I think that what's really different is the fact that we can take these lab-based capabilities in this emerging technology and the neutral atom technology really allows us to go into the field at the edge without having to take a bunch of cooling capacity, et cetera, and being able to take that out in a rugged, austere environment and actually be able to operate it at the edge where I think where decision value comes in, right? Data to decision is a lot of the area that I work in. And being able to do that with exquisite sensors and be able to actually operate that at the tactical edge where things are actually happening is, I think, the big differentiator for me.

Thank you. Grant, we'll try you again maybe for introduction and also to answer the question if you're able to hear us. No, unfortunately, still no sound. So, Tanner, maybe we'll go to you while we're figuring out the AV issue.

Yeah, thanks, Paul. So, what's changed on our end is the Brent environment has outpaced the GPS or GNSS infrastructure. Jamming, spoofing, meekening, they're no longer just edge cases that happen occasionally. What we're seeing with our customers is they're really table stakes for operating and contested operations. So that shift has forced customers to stop treating GNSS resiliency or GPS resiliency as a future requirement and really start treating it as a present procurement requirement today. And at Safran, we've seen that urgency translate into our procurement conversations. It's not just a research and development. It's not just a lab use. the other thing that's changed is the state of quantum hardware as we've talked about or as you've heard a lot today has crossed a maturity threshold where it can be integrated into the existing systems that our customers use and that's really what makes this moment real it's what makes it really exciting and that's why that's why we're excited about the partnership

thank you i'm cool i don't know if you can hear me now we can hear you now fantastic Yeah, so maybe if you could introduce yourself, and then if you didn't hear it, Grant, I could repeat the question for you.

Well, I'm Grant Shapps. I'm the former UK Defence Secretary. And whilst I was in office, Quantum was one of those subjects that went from sort of a theory, at least the way it was seen inside the Ministry of Defence, to a real here and now technology that was starting to be incorporated at the very early stages and subsequently has become very mainstream in thinking. I didn't actually hear the question, though, so by all means go ahead.

Yes, I mean, I think you partially addressed it, and the question was, as these technologies are moving now out of the lab into real-world deployment, what is driving the interest and impetus and making this kind of the moment for quantum sensing to uh to be taken out into uh real world use i think a lot of practical

examples actually uh whilst i was defense secretary i went to visit our nato troops all the nato troops in poland and i took a an aircraft full a royal air force aircraft full of the defense correspondence and on the way back i was up front in the uh in the in the cabin with the pilots at the cockpit. And we had a GPS dial as we were flying past Belarus. And I'm a pilot, so I thought this was just very fun to discuss with the pilots who weren't particularly concerned. There are other ways of navigating in the short term until it becomes outdated over a period of time. But when I went back to the cabin, the press corps were in chaos. And I thought, my goodness, They must know that we've had a GPS denial. And I asked them, what's going on? They went, we've lost our Wi-Fi. And for them, this was more significant than the GPS denial. But it does bring home to your point. The reason why quantum is now where it is and just accelerating off in terms of importance is, you know, it has the answer, the solution to things like that PNT, to the denial of GNSS and combined with everything we've seen in this European theatre with GPS denial, it is absolutely, you know, the forefront of solutions. So I think in a very, very real sense, and I mentioned that with journalists because, you know, it's the first time I think they'd ever thought about it in their lives. This has become mainstream rather than a technical

concern for people. Thank you. Yeah, I remember seeing that in the press at the time at clearly made a big impact. Chris, maybe I can ask you a question. I think one of the areas where we see a lot of interest today is in quantum radio frequency sensing. If you could talk maybe a little bit about what's driving the interest, where do you see the near-term

applications of that technology? Yeah. So most of the customers we work with are really focused on being able to see something first, be able to understand it first, and then be able to act first. And so the neutral atom technology and quantum RF enables, first of all, more sensitive solutions. And so, in other words, we're able to see certain things first. You know, secondly, we're able to make things smaller, more, you know, make them happen faster, and also, you know, more sensitive. And so we're able to now utilize these devices in configurations that we couldn't use, you know, in the past. And then we're able to basically network these all and expand upon the vision that we were able to do previously. And so we can now be able to protect our warfighter and be able to identify these signals of interest that we couldn't in a faster way that we couldn't in the past.

And Tom, as somebody who deals with mission customers on a regular basis, as you think about taking quantum sensors out and both quantum clocks, RF sensors, inertial sensors, what does a successful deployment look like from your standpoint and from the end customer standpoint?

So I think from my perspective, the swap C is critically important, right? How do we get down to smaller sizes, lower power, lower cost per unit, right? Because typically in a lot of the broader sensing community, whether it's RF or other, size is a big deal, right? So if you have a very large platform that's operating somewhere and it's promoting its mission value, But if you can take that same sort of capability and distill it down to something that's on a drone or smaller, expendable kind of thing without people being forward deployed, I think there's real value and capability there. And for me, that's one of the most exciting things about the sensing technology that really is going to allow us to have better overall situational awareness in a contested environment, but also be able to do so in a way that maybe doesn't put human life at risk.

Very good. Thank you. Tanner, I'll turn to you now. So I mentioned before in my remarks about the partnership that Inflection and Safran have established. Maybe you could talk a little bit about why Safran chose to partner with Inflection and what you're expecting to see out of this relationship in the years ahead.

Yeah, absolutely. We looked at quite a few opportunities, and our team was, and I personally was very impressed with Inflection from the first time we toured the facility in Colorado. When we had an initial demonstration of the quantum technology, it was incredibly impressive to myself personally, as well as the entire team. A couple of highlights that really stood out and helped convince us that the partnership was the right person with inflection, the right entity with inflection, and that the time was right, was the fact that the ticker optical clock is not a lab prototype. It's been demonstrated operationally in GNSS-denied environments, and that's a real meaningful bar to clear. Next is product compatibility, as we demonstrated in the Quantum Corridor. The neutral atom approach integrates directly with our White Rabbit and SecureSync environment in the platforms. As we were looking to expand our product portfolio, we were really looking to extend it with quantum-grade holdover performance and Inflection fit that model perfectly. And then finally is the organizational alignment. Both organizations share the conviction that resilient timing is foundational to our infrastructure. Inflection Quantum Sensing Portfolio is a production-ready solution that our global distribution network at Safran can both deliver and support through the life of the product.

Great. Thank you. Thank you very much. Grant, I'll turn to you now as somebody both obviously who was responsible for Britain's defense, but now also as the co-founder of an innovative defense tech company. How does the conflict in Ukraine and the emerging conflict now with Iran affect government's perspectives on adoption of technologies, new technologies like quantum, and the speed with which they're looking to deploy and to procure?

Yeah, I guess I struggled as defense secretary with the same thing that every defense secretary ever has had to worry about, which is you might commission a new project, and about 15 to 20 years later, the thing turns up in full service. I mean, it would be unimaginable in almost any other walk of life um now we have a situation in no small part because of what's been going on in mainland europe with ukraine and the speed of iteration but also now in just the last few days i mean the last 10 days what's been going on in in iran uh where governments and and um the procurement departments for departments of war ministries of defense are turning around and going we need a solution to this problem and we need it now i mean in my other uh wearing my other hat as co-founder and share a Cambridge aerospace, that happens to be effective interceptors for drones and missiles, not least because the cost of firing things like the Shaheed drone is so much lower. So you're seeing a sort of pace of adoption, which is, you know, coming down to, you know, months, certainly, but sometimes even weeks and days, and iterations, and particularly in platforms which are software-driven, which are daily and sometimes hourly in process. So I've seen, going back to the specific example of quantum, the way that there is a drive to try to miniaturize everything that's happening in order that we can get it out into the field. And, you know, rather than these being large encased cabinets in large rooms with lasers, how do you get that thing down? And I've seen a lot of procurement activity around and excitement, actually, frankly, around doing that, particularly, as I say, combined with the need to know precisely where you are in ways which aren't as abstract or aren't as specific, I should say, as the aircraft, as a jammed GPS. Yes, I'm talking about if a hypersonic missile is flying at five times Mach arc, what does that mean in terms of a very, very small outage in the location for that, just a minuscule fraction of a millionth of a second? The answer is because of the speed, it means a hell of a lot. So getting all of these technologies to actually work for the defense world as it is today, as opposed to how it will be in 15 to 20 years' time or was in 15, 20 years behind us is really the challenge now for ministries of defence, departments of war and the rest of them.

Great, thank you. Well, again, glad to be working at a company that is focused on the miniaturisation as a key element of our strategy in context of what you said. As we just have a few minutes left for this session, maybe we can kind of go to a wrap-up question. And, Tom, I'll start with you. So looking out five years, which I know in quantum is challenging, even looking 12 to 24 months ahead. But if you put your crystal ball hat on, looking out five years, which quantum sensing technologies do you feel will have the greatest operational impact?

Well, clearly, the ticker timing capability provides real world value today with existing systems. I would say the quantum RF and then the software framework around that is going to be key. You know, the amount of data that those sensors are going to produce has to be processed, managed, and disseminated to those that need to make decisions. And in my mind, I think in the next five years, it's fielding those capabilities out in a way that allows decision makers to go more rapidly with informed information that really helps determine outcomes. And in my mind, I think that's one of the keys over the next five years.

Great. Thank you. And Chris, same question to you.

Yeah, I would say from our perspective, the customers that we talk to, inertial sensing is a key piece. So, yeah, once again, the clocks and how to enable the function of these types of devices in GPS-denied locations as well as the quantum RF is the capabilities that we see and integrating them into hybrid type of scenarios. So in other words, let's say you're on a drone and being able to incorporate these RF devices onto this drone and then allow it to do its thing without, you know, being impacted.

Great. Thank you. Tanner, same question to you.

Yes. I lead the timing business line for Safran. I'm going to stick with what I know, and I will stick with the ticker optical clock. We've talked a lot about applicability for defense, and I agree the stakes could not be higher for our U.S. and our partner defense forces. what we what we haven't talked about is that the same gps dependency that defense has creates a vulnerability in in our commercial customers as well uh every data center every telecom network the entire energy grid the financial exchanges including the new york stock exchange all rely on gps timing that that has increasing vulnerability with it that's an excellent point

thank you for uh for clarifying uh and expounding on that and then uh grant a final uh comment on

this uh from you yeah well we i mean we've spoken a lot about the timing element and you know timing is like this sort of you know the most basic form of everything that happens in defense i mean in every possible way not just the hypersonic missiles but um you know gps and the rest of movies just as we've discussed this actually i i round here on something else which has been really troubling during the ukraine conflict which is the moment you put on a um a radar system uh within 10 minutes a heat see a heat-seeking missile just comes and eradicates it and i've seen this time and time again in ukraine both ways around so quantum rf is actually an opportunity to be able to sense in an entirely different way and i think this actually might be at least as important possibly even more important than the specific timing uh elements of it um and even if And even taking those two into account, there are going to be a whole range of applications that we just simply have yet to invent that are coming down the track very, very fast. Now, I know from secret level briefings that I better not go into right now that some of those have very, very significant implications for future warfare and rather defense in particular. So a lot of that is based on quantum. It's an incredibly exciting sphere of research technology. And it's now gone beyond the research into, as I say, some secret level program implementation.

Well, with that, thank you all very much. We're at time. Very much appreciate your participation and the great discussion today. Thank you very much. Thanks.

Thank you for time and for joining us for this. So we're hitting our last section of the speakers, and it's my pleasure to introduce our new CFO, Alain Hart.

Good morning, everyone, and thank you, Marcus. And I'm very pleased to be here today and meet all the analysts in the room and the one on the webcast. So while, as we said, the focus today is not to talk about inflection financials or forward-looking guidance. We will do it in details, don't worry. in details when we post q1 in may i really want to take this opportunity and share a little bit about my background why i'm standing here today in front of you joining inflection and some of the financial principles that will guide the company and i believe will enhance shareholder value in the long term so i joined the company recently you know about five six months ago, after spending more than two decades at Intel, where I held different finance leadership role across almost every aspect of the company, from process technology through CPU, graphics, wireless, GNSS, Wi-Fi development, and toward the end as the head of finance of one of Intel's largest business units, which is a mobile client platform. Following my time at Intel, I moved over to Zoox, which is an autonomous vehicle company that was acquired by Amazon in 2020. I served as Zoox CFO for five years under Amazon leadership. We operate as a separate subsidiary where Amazon executive, or as they call themselves, S-team, were our advisory board. And we're very, very close and embedded into Amazon finance and account organization. So I think similar to Matt, joining Inflection was a very easy decision for me, very easy. You can say that, you know, my 30 years of experience from big public companies like Intel and Amazon, deep knowledge of manufacturing, you know, the process technology, you know, advanced technology, AI, ML, will do the trick. But that's secondary. What really excites me about inflection is what I think, you know, Matt Pranav, Caitlin, and Paul actually did my job. Really, the broad technology leadership that the company is bringing together. We have all the ingredients that is needed to be one of the most successful companies in the quantum space. Technology leadership, execution, as of two weeks ago, were well capitalized. and I will add on it, really exceptional talent and leadership team. So now we have all those pieces in the puzzle and it's asked for us to go and unlock this huge market opportunity in compute, sensing, establish our commercial strategy, which, as we mentioned, very similar to NVIDIA and really execute, execute, and execute. it. So if we look at some of the, you know, finance principles that will guide the company in the next few months, you know, that's what I've been spending my time, you know, since I joined and for the next, you know, several months, I will start with capital allocation. We will remain very disciplined on how we deploy capital. We will prioritize investment in, you know, area that directly advance our technology leadership. It could be in logical qubits, develop software application, the sensing industry, the commercialization, swapsy, getting those product to market and targeted investment in several go-to-market initiative. So we remain disciplined, but we will make sure that our technology leadership is maintained and really extend. We're focusing on enhancing all our internal system, processes, and infrastructure that will enable the company scalable and profitable growth while we maintain a very high degree of control. One thing that guides me in my 30 years of experience is business partnership. I'm a strong believer that a strong finance organization, NSCFO, can only be a success if they have a very strong business partnership with the leadership team and the operation. So we are working in the last several months to really put this foundation in place. You know, the ability to manage and track our financial performance through clear metrics and KPI. and now that we have the balance sheet and the cash position that we have post the SPAC getting more than 550 million dollar we have the flexibility as a company to opportunistically pursue strategic initiative so if we're bringing these together overall we're going to invest in the long-term technology leadership balance it with a discipline no financials all to the goal commercialization and profitability. So I think we all get from today, and I'm confident that if we are going to deliver on the milestone that outlined by Pranav, Caitlin, and Paul, delivering 100 plus logical qubits in 2028, developing our quantum software application, commercialization, our sensing, QRF, ticker, and in our search engine, all on a unified platform. Not just that we will maintain our leadership across all the domain, ground, sky, sea, and space. We will also go into deliver significant value to our shareholders. So it was a short session, just to know everyone. We'll talk more in May, and I will hand it now to Matt for his closing remark. Thank you, everyone, and looking forward

to chat with you over lunch. Okay, well, if I played the role of appetizer to start, I guess I am the after-dinner espresso. Hopefully, you all enjoyed the main dish. On that note, does anybody know what this is? Yes, that's right, John. It's a caffeine molecule. We talk a lot about critical minerals, but this is probably the most critical molecule of all of them. It certainly is critical to me functioning every day and i see cups of coffee on nearly all of your tables um i would love to understand how this magical molecule works but the problem is if i were to try to model out the sum atomic interactions it would require a computer the size of jupiter and that's a relatively non-complex molecule but this just gets to the heart of the point that caitlin was making in her presentation, these combinatorial types of problems are the exact same problems that we are going to be solving with quantum computers. And so if you look at that as a microcosm, the caffeine molecule is a microcosm of what we can do in drug discovery, what we can do in material science. That is what gets me so excited. This is not a 50% increase in performance relative to classical standards. We're talking 10, 100, 1000x improvements, truly the next technological revolution in the framework of how carlo de perez had laid it out you've heard from most of the people up here you have not heard from dana but you heard a lot about dana from me but just to recap we really have the team so 160 physicists and engineers and growing a lot of the ip locked up and backed by many of the best invest investors on the planet built upon nobel prize winning research. We haven't talked about our board yet, but just to introduce them all to you, our board has been largely the same for the last several years. Kathy Leggo, our board chair, has been on the boards of many publicly traded semiconductor companies, probably many companies that you all cover, including Sandisk, Lamb Research, Cirrus Logic, Fairchild, the list goes on and kathy truly has been a partner to me in building inflection christina johnson who also serves on the board of cisco and historically dupont now one of the dupont spinoffs has also been the president of the ohio state university it kills me as a notre dame grad to say the ohio state university but that is how they like to refer to themselves as well as the undersecretary of energy and and many many many other great accolades david singer took my board seat when I joined full-time from Maverick, and David was my partner at Maverick for 18 years. David's been at Maverick for over 20 years, and before that was the founder and CEO of three companies that he started and took public. Don Myricks is the former CTO of the CIA and held senior roles at the NSA and has been on our board for many years. And the newest addition to our board is someone you all mostly probably know who happens to be in the audience with us today. you happen to be in in new york city today it's eric bjornhold the cfo of microchip who has been serving as our audit chair for the last few months and working alongside elon to make sure we are operating ex excellently on our audit side and working with our partners at kpmg a very very strong board and myself the least impressive background there i was at maverick for most of my career and then enjoyed inflection and i've been on the board for eight years we've got some amazing advisors as well, including General Cameron Holt and General Paul Funk from the Air Force and the Army, respectively. Really the voice of the customer, similar to the role that Dawn plays on our board. Ian Thomas, who spent a long career at Boeing, who's here with us today as well. And then Laura Thomas, no relation, but Laura spent 15 plus years in the CIA, another great voice of the customer. And the list goes on and on. I won't focus on everyone here on this list, but we have an incredibly, incredibly deep team of extremely talented people. You heard from Caitlin. You did not hear from Julie. Julie is here as well. Julie had a long career at Intel, just like Elan, working directly for Andy Grove for many years. Max Perez is in the room as well. Max, raise your hand. If you all have questions for Max, you can talk to Max at lunch. Max has been at Inflection for longer than I have and is a visionary in the quantum sensing field. And just to point out a few other folks on this list, we'll just kind of pick them at random. We have Colin Sullivan, who joined us about four or five months ago to run our UK operations. Colin spent close to three decades in the Royal Air Force running the entire Chinook division and then had several senior roles at Boeing as well as Lockheed and has been absolutely instrumental in taking us to the next level with the UK. Jim Colosimo, who's our chief engineer for QGG. Paul talked about the QGG, the quantum gravity graviometer program. This is a very large program. Jim has built his entire career on sending things to space. So Jim's going to help us get that grid of gravity gradiometer into space. And then I'll point out Carl Pendergast, our GM of our sensing group. Carl has a long history in both the precision timing as well as roles at Lockheed Martin, Ball Aerospace, etc. He knows how to build and run a sensing division, and in our case, a quantum sensing division. And then we'll pick one last one. Let's talk about Dave Kress. So Dave ran product groups at NetApp, at Nutanix, at HP, at AWS. We're really lucky to have him join recently to blaze the trail on the commercial side of things for our products. And so Dave's our VP of commercial products. But as you can see, a very long list of incredibly impressive people here. So to end where we began in my three-part simplistic framework for evaluating companies, technology, execution, and financing, hopefully you have taken away that this neutral atom platform is very powerful, and we are pointing it at a number of different applications across the quantum sensing and the quantum computing worlds, but the underlying technology is the same. it's just doing different things with the atoms. We are leading in those quantum metrics that matter, in particular in compute, with the world records for physical qubits, as well as being the only publicly traded quantum company with logical qubits, which logical qubits are the keys to the kingdom in quantum computing. From an execution perspective, I hope it came across that we are first movers and pioneers in what we believe will be the winning modality of neutral atoms and have great, robust partnerships and customers. You might heard from many of them here today. And I believe our commercialization strategy is the right one, really following in the footsteps of how NVIDIA built their business on pointing this very core, powerful platform at a number of different markets and island hopping our way, for lack of a better term. And then finally, from a financing perspective, we are incredibly well financed with a very capital-efficient business model and capital-efficient technology and are very well capitalized to accomplish this mission. And at the risk of sounding cheesy, what really does get me out of bed every morning is we are solving the world's hardest problems with the world's smallest particles. And that is why we're able to do what we can do at the limits of nature. So I hope we accomplish our goal of making this time well spent for you all. If at the very least, you got a new book recommendation and at the very, very least, you learned a new quantum dad joke. But with that said, we will call up all of the presenters, and we're happy to take any questions that you all have. And so maybe the right way to frame this is or to focus is I'll stand over here. I'll act as air traffic controller. I'll answer the questions if I feel equipped to do so. But more importantly, I'll probably air traffic control them to Pranav, Alon, Caitlin, or Paul. You guys got your chairs? Okay, cool. Perfect. We got enough. Great. Okay, so we'll take any questions from the audience, and then, of course, any questions from Julie. Maybe you can monitor the chat, and we'll take questions from the chat.

All right, so, Matt, we're going to start it off in the corner with Craig here.

Yeah, thanks for that, and to the whole team. Thanks for all the insights today. Incredibly helpful. Matt, I'll direct this to you, but you can traffic direct as you'd like, of course. So early in your remarks, you indicated the company had hundreds of customers. Can you talk about the extent to which they're engaged more on a point capability basis versus across the portfolio spectrum? And as we look at how customers are engaging now versus what's in the pipeline discussions, how should we expect that's going to evolve going forward?

Yeah, absolutely, Craig. I will take a crack and then have maybe Pranav talk about some specific compute customer use cases across software or hardware and then have Paul talk about some of our specific quantum sensing customers. And then there's also sensor customers that we have sold our core capabilities to, that quantum core. It in some ways depends on how you define customer in that you could define the U.S. Department of War or the U.K. Ministry of Defense as one customer. And in that case, you will see effectively all of our different products being deployed to those larger defined customer bases. But if you were to narrow that down into maybe more granular customers, you are seeing situations where they'll start with a quantum clock, let's say, or quantum software, and those are, or even in some cases, quantum components. And those act as the tip of the spear to get in the door and then start to expand the different types of products that we can sell to them. and maybe I'll give Pranav the opportunity to talk about some specifics where maybe software acted as the tip of the spear, and then we could expand from there, and then, Paul, we can talk about some sensing applications too.

Sure, I'll give a few examples that come to my mind. A lot of the world's leading groups studying quantum, whether it's sensing or computing, use inflections, neutral atom core, and maybe you can flash that up, and those academic groups then start to climb up their own productization journeys in oftentimes partnership with us. So I mentioned that the Institute for Molecular Sciences in Japan has purchased a QPU from us. And that's one of the relationships where it started with this core technology that Inflection delivers to a variety of university labs, to national institutes, to defense organizations, to enterprise, et cetera. And that was an example of climbing up from delivering these cells to going up to delivering an entire quantum computer, a QPU. Another example that I'll give you is this work that we're doing with the partnership with NVIDIA. So that started with SuperStack, our quantum compiler platform. And since as early as, I want to say, 2023, we collaborated with them on this data center application called SuperCheck. That was before NVIDIA had truly leaned in on quantum computing. And that was the seed that spilled over from the quantum software land into quantum computing. And now our machine is going to be on site at NVIDIA's booth next week at GDC. The last example I'll give, and then maybe pass it over to Paul, is with the U.S. Army. It's another example where we've been able to put our software applications first because it works for both classical compute and classical sensors. But that has ignited, unsurprisingly, a lot of conversations in terms of now Army's bought in on, you're already helping us see where the software stack is going to integrate sensors, what new sensors are coming, and naturally we've been able to make the right introductions to the quantum rf sensor the ticker timekeeping the inertial sensor etc so there's it's quite dramatic how much conversations in one part of our company lead to business in the other part of the company in fact both parts of

the company yeah it's a great question i'll give you uh maybe three uh examples um and first is to follow-on from Pranav's comments on the Moonshot QPU. We sold, as Pranav rightly says, initially core components to the institute, then we sold them the QPU, and actually subsequently to that, we've now sold them software as well. These become very sticky, very long-term relationships over time as customers utilize and get greater utility out of multiple pieces of our offering stack. I think the second example I'll give you is in QRF, quantum radio frequency sensing. We're now actually, we are incorporating together both the core quantum RF receiver and ticker. And so by adding ticker to the quantum RF receiver, we're able to improve the performance of the system. We've just shipped one actually just about a week ago to a customer, and it was the first one where actually it's the two products together in an integrated solution. I think I just thought of a fourth one. So the third one I'll give you is the point that Tanner made in the sensing panel where a key part of the relationship with Safran is not just Safran reselling Ticker into their install base, which certainly is fantastic as an accelerant, but actually we're bringing together Ticker as the time and frequency reference with the Safran White Rabbit solution for extremely precise timing synchronization. So that was the example of the work that we did and that Max pioneered at Quantum Corridor. Now we can bring a solution to market that does something completely new that was never possible before by integrating these pieces of the solution stack together. And then I think the fourth point that I'll make is, I think, if you think about PNT more generally, which is this area of operation in GPS-denied and contested environments, what is PNT? It's position, navigation, and timing. Well, to do that, you need timing, obviously. It's in the name. But you also need the navigational component, which is inertial sensing. So I think we have kind of a 1 plus 1 equals 3 equation with a number of the pieces of our solution set.

I think Paul said a really important thing there in the example of Ticker being a component of QRF and improving the performance. Some of these sensing products are, as I pointed out at the beginning, some of the basic building blocks of quantum computers. But they're also the basic building blocks of each other. And so quantum computers require precision timing to synchronize the lasers. And so these clocks are building blocks of many, many different types of products. Can I go with Quinn? No, sure. No. Here we go.

Okay, I won. John McPeak, Rosenblatt Securities. Thanks for doing the day. It's been very helpful, guys. This is mostly for Pranav, but possibly Caitlin as well. The gate speed for the atomic modalities is critiqued by the modalities that have bad error rates, but high gate speeds, could you talk a little bit about your all-to-all connectivity and how you can parallelize the circuits a little bit and get around that? Thank you. Yeah, this is a great

question. And maybe I'll expand it just for the audience's awareness. So superconducting quantum computers, like the ones being built by IBM, Google, et cetera, they have the advantage of of running at about one megahertz clock speeds. So pretty darn fast. The neutral atom approach is roughly, let's call it about 10 kilohertz. So indeed 100X slower. So that initially at a first glance sounds like a barrier, right? But what it's really important that John has teased out there is that every step in a neutral atom quantum computer is much more powerful than every step on a superconducting computer. And the best way to think about this is ultimately we want to solve customer problems. And the way that we solve a problem is we take a number of steps, and each step takes a certain amount of time. So on neutral atom computers, the number of steps that we have to take is much smaller. Every individual step is 100x slower, but the number of steps we need can be as much as 1000x less. And the reason for this, you mentioned the word all-to-all connectivity, is that I can make this qubit interact with this qubit on a neutral atom quantum computer in a constant step on a other modality like superconducting that operation itself might take 300 steps and so on balance these things roughly cancel out and when it actually comes to time to solution neutral atoms end up stacking pretty much equivalently to superconducting on actually solving real problems and this is probably familiar to a lot of us in just the consumer world, too. When you went to Best Buy in circa 2005, there was sort of a race of, oh, well, you have a 2.4 gigahertz processor. I've got a 3.5 gigahertz overclocked Pentium 4 or whatever. And now you've seen the race to sort of go back and say, well, we can do more with one gigahertz, less power consumptive, but each step and each step does way more compute. It's deeply related to the chips analyst tiers on the RISC versus CISC trade-off that happened in the 90s. So that's the deep answer, but the short is each step is slower, but the number of steps that we need to take to solve a real problem is dramatically less on neutral atoms because we can make every qubit talk to every qubit in a constant number of steps.

I got it.

This is Tyler Anderson from Craig Hallam. I was looking into the ticker C. What kind of cost reduction do you expect to get from that? And when we think about data center applications, is this able to address scale across and then at what point do we get into like between data centers to do the network synchronization and then when we think about intra data center is this is this more of a scale out or is this like on every top of rack just thinking about that moving forward

sure I can take a quick crack and then Paul you can go so roughly speaking and I'll stick as opposed to bill of materials and maybe sticker price so we can have that conversation. So right now we sell our clocks for about the current version of ticker for about $225,000 a pop. Maybe we'll use rough numbers, $200,000. And as Paul mentioned, the goal is to get on price parity with the existing precision clock technology, which let's call that roughly $100,000 or so. And our goal is to get to that $100,000 point as fast as possible. There are a couple of ways to do that. One of them is to continue to integrate the different components inside the bill of materials. And one of those, again, goes back to what Paul was talking about, is about photonic integrated circuits. So taking dedicated laser systems, integrating them down to silicon, and therefore you can drive the bill of materials down materially. The other is volume. right? And so selling more of these clocks will help drive down the price as well. And that's really been the root of the strategy, which is, for lack of a better term, go to where the dogs are eating the dog food now, which is in the national security world, use that to drive up some of the volumes, drive down the cost to address that broader commercial opportunity that Tanner teased during the Q&A, because this is not just a national security problem. This is ubiquitous across commercial entities being reliant upon GPS for the disciplining of their time. So I didn't answer your specific question as to what I expect from ticker C, but maybe that's a framework to how to think about it, and we're working towards that price parity as fast as possible. Paul,

you want to add anything to that? No, I think that's a good answer to the question. To the subsequent follow-up part of your question, in terms of inter-data center synchronization, that's essentially the first step we took along that path was back in, I believe it was November of last year with a demonstration with Quantum Corridor, which we did utilizing the White Rabbit protocol of the White Rabbit system from Safran across that urban fiber network. So expect to see more of that as we look to roll out the Safran partnership kind of in earnest in the coming months. I think within the data center, the question of is it top of rack, top of spline, actually Max has spent a lot of time thinking about that. So maybe during the break, you could follow up with him. But I think shortly, simply put, we see the ticker family of ticker prime, ticker C sitting at the master node within a data center. And then ultimately, as we work towards, as I talked about, longer term, that ticker blade, the card scale implementation, then ultimately the goal there would be to replace CSACs and MACs at a much lower cost per unit, but at much higher precision.

All right, Matt. So we have a question from back here in the audience. All right, then we'll do Quinn.

Kingsley Crane at Canaccord. And the question is on one of focus. So across the quantum space, think of the grand prize in quantum compute, building, maintaining, extending, and leading quantum compute, reaching 100, 200-plus logical qubits, and then probably various other prizes in areas like sensing, networking, CML. that could still be quite large since it could be a $30 billion market. So the question is how you think about balancing those operationally and focus internally.

Yeah, it's a good question, one that I think about a lot. One really important thing to mention, though, is focusing on the interconnectivity of all the different products. So in many ways, as we're continuing to commercialize and capture the opportunities, in clocks and sensors more broadly, that is directly adding to the speed to which we get to the 100 logical qubit level, the 1000 logical qubit level, because the underlying technology is all the same. I think your question is more along the lines of, when you get to the beginnings of that crown jewel of useful quantum computing, what do you do with the sensing opportunity? The way I think about this, and I hearken back to the way NVIDIA built their business, is they didn't stop selling GPUs to the gaming market. They didn't stop selling GPUs to the crypto mining market or the physics market, and they continued to service all the way up to the large language model market. That's my vision as to how we will continue to operate as inflection. These market opportunities in sensing are massive, as is the market opportunity in computing. And the good news is the underlying technology is very, very similar. So not exactly an answer to you, Kingsley, but I think that's generally how I think about this.

Quinn. Okay. Quinn Walton with Natum, thank you for the day. I guess I want to start with a high-level question. Ilan, you had a short presentation, but I think the message was clear. You want to have financial discipline as you come to the public markets. My question is for Matt. Many of your public peers, once they got public, put their foot on the gas, accessed capital markets, used their currency to acquire companies, and may have accelerated near-term losses in the idea that this is a land grab, this is very early on in what could be a very large opportunity. And so how do you balance financial discipline in the near term versus going out, getting bigger, whether it's through M&A, whether it's just through additional hiring and trying to maximize your opportunity to become a leader in not only compute, but also quantum sensing. And then I've got a technical follow-up.

Okay, great. No, also a very good question, Quinn. So the good news for neutral atoms is it is a very capital-efficient modality, and that's compounded by the strategy which links these products together in a deeply integrated way. And I'll repeat it again, but the underlying componentry of the clock is just very, very similar to the computer and everything in between. So as we are, I'll stop beating that dead horse, but there's an immense amount of technology as well as financial leverage in that model because of the underlying neutral item modality. So that's one thing we have going for us. Another thing that we have going for us is one of the reasons why I felt very comfortable raising the amount of capital that we did is because I trust myself and our team not to become drunken sailors, for lack of a better term. We are going to, with my investor background look at capital deployment with a ROI mindset. And so to the extent that we do see great ROI opportunities, we will accelerate capital deployment into R&D or go to market to the extent that we think that earns a long-term positive ROI for our investors. The last thing I'll comment on is acquisitions. We will, with a public currency, have more flexibility to do acquisitions. That said, I, again, with an investor background, go into acquisitions with a pretty big degree of skepticism. Most of them just don't work. And so we will scrutinize heavily any acquisitions we make. There is one area where acquisitions can be successful, and that is technology tuck-ins that directly accelerate your roadmap. So if we were to do them um hypothetically that would be where we would be open to doing it something that would directly accelerate our cost down for clock something that would directly accelerate quantum rf or inertial sensing and so i think that would be how i would think about the acquisition side of things but couched in a great deal of skepticism from the beginning and then on um just on how should you think about you know the burn rate going forward i think is sort of at the heart your question um i'll just refer back to my maybe not not the best language of drunken sailors but i do think that we will not materially accelerate our our burn rate um so i wouldn't be expecting that to occur and you know a year from now we may make an roi decision that might be different than

that but um that's the plan for now perfect and then i guess this may be for pranav um you highlighted quantum error correction is sort of key to getting to the 30 and then the 100 logical qubits. But you haven't sort of spoken a lot about what kind of error codes you're looking at. And maybe just spend a minute. You've got the highest number of physical qubits of any modality on the market today. You've got very high fidelity rates. You've got all-to-all connectivity, I think, which all favor efficient quantum error correction. Can you talk about what codes you're looking at using and then what's the biggest challenge to implementing the quantum error correction on a neutral atom platform? Is there one specific hurdle you think is the highest to overcome?

Thank you. So let me tag team this with Caitlin too and maybe I'll lean on you to explain some of the co-design work that we've done, like the material science. but to step back one of the immense opportunities on neutral atom platforms is it's a bit of a mouthful but it's called qldpc and in fact all of you are using the last four letters ldpc right now if you're on the wi-fi here so ldpc stands for low density parity check it's mouthful but it's basically a very efficient way of taking the signals that the wi-fi routers up there are sending and converting it to pristine data on your laptop even though that wi-fi signal is bouncing off the wall, sometimes getting dropped, etc., you're still getting a very clear signal, hopefully, on your laptops. And in fact, it turns out that you can take the same error correction technique in routers and bring them to quantum computers. So I had my penultimate slide in my first segment was this quick survey of just things I didn't have time to tell you about. But one of them was we launched a software library called QLDPC, and that was in collaboration with Jake and Morgan Chase, which as some of you may know has one of the best quantum computing teams out there, they and us are deeply interested in, can we take the ratio, the cost of building logical qubits down from about three years back, the conventional wisdom was that to get one logical qubit, you would need a thousand physical qubits. Today, we see a path thanks to the software library that we've collaborated on with Jake and Morgan Chase and co-design with Neutral Atom Quantum Computers to bring that ratio down to 24 to 1 instead of 1,000 to 1. So to the crux of your question, the neutral atom opportunity is quite unique because we can make every qubit talk to every other qubit and have highly parallel interactions. That means our price that we pay per logical qubit can fall dramatically lower than what we've done. And that paper that I showed in one of my slides on efficient error correction, that was a case where we actually got down to four physical qubits per logical qubit with quantum error correction distances. In practice, we're going to need to scale up to more like 20, 30 physical qubits, but it is a very unique capability that we have, and a reason why inflection has logical qubits, I know Matt mentioned this, but we're the only public company that has logical qubits, and it's an advantage of the neutral approach. We're at 12 now, expecting 30 later this year. And maybe the last piece to pass it over to Caitlin on is how our application team and our software effort is then co-designing applications on top of that.

Yeah, so maybe one, I'll emphasize a couple of things. So Pranav hit on the main part of your point, but within our computing division, we have robust capabilities all across that stack. So obviously we have an amazing hardware team. We have an amazing compiler team. We have an amazing applications team. And each of them have their own objectives, their own mission. But really, when it comes down to these big logical qubit demonstrations that we've been pushing out recently, we come together as a single unified team. We have a unified vision about what it is that we're trying to achieve. And we think that that is absolutely crucial for doing these kinds of demonstrations. And it does involve that exact question that you just asked, which is identifying what are the codes that we want to target for this application? What is the best we can do right now? how should we exactly make the fine-grained details in terms of how we're going to progress right here? And I just kind of want to emphasize that one amazing thing about that co-design effort is that our team is large enough to be mature. We have mature practices. We have extremely experienced physicists and engineers and scientists on our team. But at the same time, we are small enough and agile enough that we can really find the best course and go there. We're not so rigid that we can't look at that value and say, there's more value over there. Let's go that direction.

All right. So we have a few speed dating questions from our online folks. I'm going to smash the two together. So during a previous road show, Flexion talked about having a goal for bookings for 2025, talked about $50 million. Can you update on what happened in 2025? They're basically saying we're tiptoeing around the no forward-looking comments. They want to know about the past. And can you split out sensing versus computing?

So just to define terms, we had said during our roadshow that we had booked or were awarded $50 million, just to be clear. Not just a bookings number, but also booked and awarded. And the only reason I differentiate there is because often when you're dealing with the U.S. government, there is a lag between getting awarded the business and then the contract being signed. And in some cases that can be a while, especially when the government is shut down. Hopefully that will be truncated going forward. So I don't have an update to that number, Marcus, but maybe I will comment on one thing. And that is within that bookings number. And for those of you who've covered the semiconductor industry, you can understand there's a lot that goes into bookings, right? And so they're not just one-year bookings. They can be multi-year bookings as well. So just keep that in mind that there's a whole kind of hodgepods of different types of lengths of contracts that are within that number. And we put that number out in September or so. But there will be more updates on our financials at the earnings call. What was the second question?

They wanted a color between the sensing and compute size.

Okay, sure. So I think the way to think about this historically has been sensing has been more than 50% of our revenue. And this isn't a precise number, but I think it wouldn't be terribly off the mark to say it was two-thirds or about that number. And therefore, inherently, computing has been sort of one-third or below that. And that has been the historical levels. going forward, I would anticipate it to probably be somewhere in that range, but with some pretty wide error bars, particularly around compute, because these sales are pretty lumpy, right? When you book a big computing sale, that will be a large amount of bookings, which will then change the revenue mix over time. So I think it's probably not a bad thought to think about it in that historical, but also potentially inverse. That's a very wide bar there. So maybe think about 50 50 going forward for the near future and then as we get to commercial usefulness of quantum computers i would expect that mix to flip to you know be majority quantum computers and possibly you know uh possibly super majority quantum computers um over time as we get to 100 logical qubits and beyond so there's a maybe high level framework on how to how to think about things

that sounds perfect and we can all smell the food so i think this would be a great time

to wrap it up okay all right well thank you for those questions we unfortunately have a little bit of a strange um layout for lunch so uh i think people would just sit at your sit at your area and eat and maybe i and the team will wander around and mingle and we can have conversations that way, or we can stand and eat, or we'll figure it out. But so this is, we're going to wrap up Q&A and move on to lunch, which means we'll wrap up the analyst day. And so I'll just say what I said before. Thank you all for being here. Thank you for your interest in inflection. Thank you for all who are tuning in on the webcast. And we're really, really looking forward to working with you all over the coming years and hopefully decades. So thanks and enjoy lunch.

Okay, thanks.