IonQ, Inc. (IONQ) Earnings Call Transcript & Summary

Thank you for joining us for today's live webinar, IonQ's Full-Stack Quantum Innovation. We really appreciate you taking the time out of your day to listen to us live. I'm Nikhil Dhingra, Director of Product Marketing here at IonQ, and I'll be kicking this off. Before we get started, just a few housekeeping items. Time permitting, we're going to take a few questions at the end of the show. [Operator Instructions] At the end of the session, you'll get an automatic survey sent to your email. It's a short 2- to 3-minute survey, and we greatly appreciate your feedback on it. A recording of this webinar will be available after the show. Registrants will be notified when it's available. We'll get started here. So again, welcome, and thank you for joining us for today's live webinar, IonQ's Full-Stack Quantum Innovation. Today, you'll hear about key initiatives around Hybrid Quantum Computing, Customer Applications and more. And of course, you'll hear about how all of these initiatives drive IonQ's core pillars of performance, scale and enterprise-grade solutions. We've got a great lineup of IonQ technical and business leadership and expertise here today. We've got Coleman Collins, Director of Product Management, who runs many of our key initiatives around Software and Hybrid Quantum technology. We've got Dr. Masako Yamada, Director of Applications Development, who works with external customers, partners and collaborators to develop Hybrid Quantum applications that address their technical and business needs. We've got Dr. Bjorn Flatt, Head of Emerging Technologies, who drives much of our long-term technology development for future generation products. And at the core of his work is scale, and you'll hear much more about that today. You'll also hear from our Chief Marketing Officer, Margaret Arakawa, who leads our marketing, comms and product marketing functions and partners closely with our go-to-market team. So I'm not going to take up your valuable time to read this slide aloud. However, we advise you to please review our website and review similar cautionary notes, especially in advance of making any investment decision in IonQ. And with that, I will pass it over to our Chief Marketing Officer, Margaret Arakawa.

Good morning, and good afternoon and good evening, everyone. Again, Margaret Arakawa, Chief Marketing Officer for IonQ. I joined the company a little over a year ago. And one of the things that really compelled me to join the company is this mission statement, "Building the world's best quantum computers to solve the world's most complex and impactful problems." I've been working in technology and classical compute for -- I spent 20 years at Microsoft as well as different companies in high-tech. And I found over the years, it does make a difference that you have a mission and a goal and a North Star. And everyone at IonQ is dedicated to make these incredible quantum computers and the networking technologies that we build, so that they solve things that will change the lives of people in the world. I'd like to go to this next slide to talk about who have we been working with in the last years. So IonQ was founded on technology that Dr. Chris Monroe had actually did a groundbreaking work on in 1995. We were public in 2021. And since then, you can see this array of customers and partners and ecosystems that have adopted and partnered with IonQ and really jump into quantum computing. I did want to highlight on that first row, very exciting news we had recently. We were at Supercomputing '24, and we were able to show an incredible demo with NVIDIA to showcase some of the work that you'll hear later that Coleman Collins will talk about, as far as how do you think about quantum and classical computing and how do they work together? Working with a partner like NVIDIA was incredible for us to showcase that. The other 3 logos I wanted to point out, Azure, Amazon Braket and Google Cloud. We're still the only company in quantum that actually is available on these 3 largest clouds in the world. And what's really incredible is today, actually at reinvent for AWS, Amazon just announced at their session that IonQ was the one quantum computing company available on Amazon Braket that had the highest fidelity and performance of all the quantum computing platforms they support. So very exciting news out of that cloud company. As you can see, you can read the rest of them, but I'm going to actually dig in and showcase some of the things that we just announced as far as customers and partners. We just announced with our work with Qubitekk, we signed a definitive agreement to acquire the assets of Qubitekk. Qubitekk is a networking company. So if you think about our company as a quantum computing company from the hardware, the software and the applications we work on, Qubitekk is actually extending the reach of where IonQ has been focused. So we did announce that we're incredibly excited to bring in Qubitekk's commercial operations, their customer base and their employees to IonQ. With that acquisition comes 118 granted quantum networking patents. So really incredible work of Qubitekk and something that we can build on as far as IonQ's foray and expansion into networking. Another announcement that we just recently made was a collaboration with AstraZeneca. AstraZeneca is a $40 billion-plus very, very well-known pharmaceutical company based in Sweden. We are building a Quantum Application Development Center in collaboration with AstraZeneca. We'll be able to leverage the power of our quantum experts and their world-class scientists to develop applications in the BioVentureHub and really focus on how do you work with Quantum and Pharmaceuticals. How do you accelerate the pace of drug discovery. Another announcement we recently had was with ANSYS. ANSYS is one of the largest companies in the world working on computer-aided design and engineering. We have an agreement now in partnership to work with them and seamlessly integrate with ANSYS' software and IonQ computers. When you think about ANSYS, you think about all of the high-performance computing that's required to these incredibly difficult complex computer-aided engineering designs. And we're working with them. So HPC is married to quantum computing to accelerate design and manufacturing. Another really incredible announcement we had was a $54.5 million contract award with the Air Force Research Lab. This particular contract also actually showcased and expanded our work with quantum networking. This is one of the largest quantum contracts known in the United States. And not only are we working on deployability and fieldability of quantum computers, we're working on how do we actually work with interoperability with other quantum computing modalities as well as how do we work with quantum computing and drones. How do you scale similar to how we built the Internet. When you had computers just in a room, how did you network them within the room and how did you actually build the quantum Internet -- sorry, the classical Internet. We're working with the Air Force Research Lab on the quantum Internet and how we scale. And then finally, one of the biggest incredible partnerships we signed with the University of Maryland, a $9 million partnership to drive quantum innovation. I did want to also just step back and talk about our differentiators. IonQ focuses on trapped ion technology. And that is a differentiator because we start out with perfect atomic qubits with long coherence, high fidelity. We are modular, and we really focus on practical time to solution, which you will hear later from Masako and the other speakers. Again, I wanted to talk about patents because with Qubitekk's acquisition, we're actually going to have a total of over 600 U.S. and international issued and pending patents with the acquisition of Qubitekk. So when you think about a company of our size, having that patent portfolio is quite incredible. The next slide actually goes over some of the details of where we actually strengthened our technical moat in performance, error handling and scale. You can read more about that in our blogs as well as in our press release. I wanted to land on this next slide, which just says what Nikhil just talked about. Something we really focus on is how do we balance performance, scale and enterprise-grade solutions. You'll hear more about that. And in the next slide, talk about really what our journey was to get to that point. IonQ is on the fastest path to enterprise-grade quantum computing. We started out prelab. We started out experimental. We went through academic demonstrations, and we got to an R&D situation where we expanded our computing capabilities, our technological road map, and we are now at #5. We're proving out business value, deployability, interoperability and enterprise-grade. So with that, I'm going to hand it over to Coleman, who will actually talk about some of the great advancements we have in interoperability.

Yes. Thanks, Margaret. I'm Coleman, Director of Product Management here at IonQ. And like Margaret mentioned, one of the ways to think about enterprise -- this enterprise-grade pillar is as a pillar that's all about taking our high-performance, highly scalable systems and helping them become useful commercial objects as well as scientifically and technically interesting devices. And the way we do that is, by starting with the best qubit technology, building it into the best hardware. And then building or integrating all of the other pieces up the full quantum stack to enable everything else we need to drive meaningful commercial value at production scale; reliability, resilience, usability and application-oriented performance. So to walk you through this picture on screen briefly. The hardware is still at the core. It's still the most significant bit for IonQ's continued customer success and commercial success. But then on top of that, you need a robust software suite and access platform. This runs on the computer as well as around the computer. And we build a lot of the software in-house, but we also believe and support a broad quantum ecosystem of open source and third-party access platforms and software packages, including all of the major cloud providers and what they have to offer. Now all that software is there to enable this top layer, meaningful commercial application development and deployment, which again happens in-house via IonQ's stellar application development team and often also with amazing partners like the ones listed here and sometimes even through people outside of the building. We regularly see application breakthroughs coming from places we didn't expect because we have this robust access platform. And so for the rest of the session, my colleagues and I are going to walk through some of our latest updates in each of these 3 layers. Starting with the middle, because it's -- like Margaret mentioned, this word interoperability, it's really the place where we are able to bring these 2 sides, these other 2 layers of the stack together to create useful applications and meaningful progress. So if we'll jump to the next slide. This hybrid story is really all about tying these 2 pieces together, right? I think a lot of the ecosystem and messaging around quantum is still treating quantum processors as a sort of stand-alone devices often a special corner to kind of tinker or play with or learn about. But the long-term vision here is commercial advantage. And if we make the bar for taking advantage of a quantum computer, totally reinventing your existing classical infrastructure, the way you solve problems in the world today with computers, the places where quantum computers are likely to provide advantage and the people who will be able to help are going to shrink dramatically. And this is really what we mean when we say hybrid. It's more of a road map theme, a collection of capabilities than any one individual feature or option that we provide. And they're all about bringing quantum computing into the existing production classical computing environment as a useful tool, which primarily means integration operability with classical computing. So we're going to start with this new cluster of capabilities that we're launching under the banner of IonQ Hybrid Services. These are a collection of capabilities that span all the way from the cloud, all the way down to the hardware to help enable the kinds of quantum classical integration that we believe are going to be key to unlocking commercial value at scale by allowing us to effectively take advantage of the things that aren't the quantum computer to provide real value. And if you'll go to the next slide, to put all of that a bit more simply, this is what we're trying to do in the long term. Quantum should just be part of the production value stack, another way of potentially many to get the best results, simple as. We want the capabilities of quantum computing, the value it unlocks, the new things that will let people do to be incredibly exciting, world changing, right? But we want the act of actually using a quantum computer to be effectively boring; as simple as it is to take advantage of a GPU to accelerate parts of a workload today. And this is starting to be a story that's more prevalent, I think, in the industry. But a lot of folks, including some hardware vendors, don't really have a sense of what it means to actually do that, the full path to getting there to get all the way to production -- sorry, this is a bit of a mouthful, a production quantum accelerated application. There we go. And so, I'd like to talk about how IonQ thinks about that on the next slide. How do we do that in detail? We like to think of it here at IonQ as a virtuous cycle of how quantum computers can accelerate classical workloads and workflows sort of on this top half. And on the other side, how classical can actually enhance the quantum computers themselves through all sorts of techniques that enhance performance, speed, reliability and usability through software alone without any change to the physics, the optum mechanics, any of these things. And for now, for us, this exists mostly as 2 different road map thrusts. Here at the top, we're driving towards improving the end user programming and accelerating at the application layer. Whereas at the bottom here, we're driving towards improving and ultimately virtualizing the qubits and the QPUs via photonic interconnects, advanced error mitigation and error correction, all of which require the ability to inject classical logic into a given quantum program at certain points. So to actually do a quantum error correction routine, you need to detect the errors, which is a thing that requires a little bit of classical computation. They need to be able to take different branches on your quantum gates to then actually correct for those errors, all while the computation is still running while your quantum information is still coherent. And that requires new capabilities sort of classical enhancing quantum. And for this third piece, we do all this in what we call an application-driven way. We believe it's necessary to start with the end in mind and enable the researchers of today to help us understand what we need to -- for the production applications of tomorrow. But to be clear, we're already starting to see the lines blur between these 2 sides between kind of the quantum accelerates classical and the classical enhances quantum. And we think in the long run, these 2 will get slowly more and more like one thing as these systems become more tightly integrated. And so that's sort of how we think about it, but how do we measure ourselves? What's the scorecard here? So if we go to the next slide, if we've done our job right, the benefits you'll see as an application researcher or an end user or even just someone who's kind of further down the stream of a quantum advantage application, say you're a chemist or a material scientist and you're just using this thing and you're not actually writing the circuits yourselves, they can be described in kind of 3 big buckets. And if you look at these buckets, they're very similar to the buckets you'd see when describing kind of any high-performance computing platform, and that's the point. These -- that's the application and value-driven focus we want to bring to the table here. So to give you an example, individual gate speed is important, and it's a key factor in overall algorithmic time to solution, but that overall algorithmic time to solution is the thing we actually care about. That's the thing that creates real commercial value. Similarly, qubit performance is incredibly important and a key driver of algorithmic performance, but there are a lot of other things we can do to enhance algorithmic performance, and we want to be pushing forward on those just as well, just while we are also pushing forward on qubit performance. And of course, usability is also critical. Because like in the other 2 cases, we want to move away from raw qubits and circuits and towards value. And the way we do that is by making all the stuff around running a circuit as efficient and simple as possible. We want our applications researchers and other applications researchers to be doing their best work on IonQ systems, not fighting with the infrastructure, fighting with native gate mapping or anything like that, unless, of course, that's where your research is. We're not removing low-level access. To be clear, we're creating tools and abstraction layers that let users choose what parts of the stack they want to get deep on and what parts they want to just work. In terms of what we're actually launching this year, this is the stuff that we talked about a little bit in the press release that I think came out yesterday as well as we'll have some blog posts later this week, I believe, about this in more detail. If you'll go to the next slide. Here's what we're focusing on this year to start kind of moving the story forward. It's the first of a lot of exciting stuff to come. So first, we've introduced what we're calling hosted workflows, quantum functions and hybrid solvers, a collection of new interaction models that work together to enable these hybrid solutions and make them effective and simple, whether that's in the cloud, on-prem or some combination by packaging and simplifying the classical logic infrastructure that allows those workloads to run. Second, we're introducing a scheduling capability called Sessions, making sure that time to solution is maximized and resource utilization is also maximized is critical in the long term and in the short term. And so, Sessions really combines the best of on-demand scheduling and reservation-based scheduling to maximize both of those things. And third, we're rolling what we're calling an accelerated on-QPU run time as part of the IonQ Quantum OS, and I'll talk about that more in a minute. And all of these, to be clear, are built flexibly and can work as enablers for partner capabilities like the NVIDIA CUDA-Q kernels that we took advantage of at the demo that Margaret mentioned at SC. Amazon Braket's Hybrid Jobs like have been actually demoed just this week at re:Invent as part of some capabilities that they're working on with Q-CTRL or Azure Quantum Sessions functionality. Or in addition to that, directly through the IonQ Quantum Cloud or via our new IonQ SDK. And to be clear, again, this is really just the beginning. We have a series of additional enhancements planned to the things we just rolled out as well as additional new features throughout the Forte Enterprise and Tempo time lines to get us to this goal of large commercial advantage applications running in production. I could talk about the kind of individual things that get us there, the bullets on this slide, but I'd rather talk about what this all builds to, which is a world where future developers don't care about physical qubits or physical machines and simply pass application inputs to production applications and get a quantum accelerated result. The value of these capabilities now is that it makes development easier. But in the long term, the value is that anyone who needs this power can leverage it regardless of quantum experience. And there's quite a bit of work to go here, but it all starts with the foundation we're laying right now today, allowing application developers to describe and schedule workloads in a way that creates these packaged portable descriptions of the computational goal, not just the quantum circuit. I mentioned I wanted to dig in a little bit on the IonQ Quantum OS, which is what we're going to do next. So like I mentioned, one of the places we've been investing heavily in the past year or so is in this major revamp of the IonQ Quantum OS ahead of IonQ Forte Enterprise moving -- or sorry, IonQ Forte moving into on-demand access and our first Forte Enterprise systems coming online very soon. Because Forte Enterprise has sort of always been our first system that was really supposed to live up to this enterprise-grade pillar. And for that feature in every respect, it needed a control fabric worthy of that capability. If we'll go to the next slide, even calling it an OS, while a useful metaphor kind of sells this thing short. It's really a platform that enables the continued rapid evolution of our R&D road map, our multi-QPU future. It includes a robust error mitigation suite, a variety of hybrid development -- and supports a variety of hybrid development integration models and so much more in addition to, if that wasn't enough, keeping our production systems running with high performance and availability. There are a ton of other amazing enhancements the team has made in getting the OS sort of ready for Forte Enterprise ready for this enterprise-grade future that are described in more detail in the blog that's coming out later this week. So right now, I'm just going to focus on the kind of end-to-end application-oriented improvements we've been able to accomplish. These include some major speedups in what we call classical overhead, the time around when the ions are actually doing computation. And when combined with some additional clever scheduling and workload management techniques, these have translated to a speed up of something like kind of 2x half the time in most cases, and in some extreme cases, up to 46x. This is taking something that took a little over an hour down to a few minutes, all while keeping algorithmic quality steady or in some cases, even improving it by these advanced error mitigation techniques. And if you'll go to the next slide, these enhancements are already running on IonQ Forte and will be included at launch in all Forte Enterprise systems. They help take what is already an incredibly high-performance system and make it ideal for the kinds of hybrid workloads and deployment models we expect it to enable, whether that's via our Cloud and Hybrid Services suite or integrated into a customer data center architecture. And like I said before, this really is just the beginning of the story. We're building a platform for the future here. And as excited as I am about what we've accomplished, I'm even more excited about the work ahead. And I think with that, I'm going to hand it off to my colleague, Masako, to talk about what our amazing applications team and partners have already been able to do with some of these capabilities on IonQ Forte. Thanks.

Thanks, Coleman. My name is Masako Yamada, and I'm the Applications -- Director of Applications Development here at IonQ. Going back to this full-stack chart, I'd like to direct your attention to the topmost bin, which is highlighted in orange. And that is where our team operates. Our applications team works very closely with enterprises in the private, public, and government sectors. And we work very closely with our partners to develop quantum -- hybrid quantum algorithms that address both their technical and business needs. We are a team of PhD researchers, scientists and engineers who span all STEM disciplines. And our 3 main focus areas are chemistry modeling, optimization, and AI and machine learning. On the next slide, I'd like to talk about some of our partners and some of the works that we've done with them. The first 3 examples we've actually talked about already in previous webinars. So I would just love for you to look at them and take a note that we have talked about these examples before. Today, I'd like to talk about the latter 2 examples with Oak Ridge National Lab and ANSYS. And in particular, I'd like to talk about a new hybrid optimization algorithm that we've been developing that is applicable to these and all optimization algorithms. On the next slide, I'd like to talk about 2 -- actually, on the next 2 slides, I'd like to talk about a novel algorithm that we've been developing with our partners, including with Oak Ridge National Lab. The algorithm is called QITE, Quantum Imaginary Time Evolution, and this is an optimization algorithm that really works well with contemporary quantum computers. First of all, I'd like to step back a little bit and talk about optimization and how these problems or the set of optimization problems can be applied towards many industries, including cargo loading, flight gate assignment, energy grid optimization. And in this case, we are also talking about optimizations for computer-aided engineering. And one of the things that I'd like you to envision is optimization as a landscape. Often, we talk about optimization in the context of a landscape that has hills and valleys. And really, the problem that we're trying to solve is if we have a landscape of hills and valleys, mountains, divots, how do I get to the bottommost point of that landscape? It's a little bit counterintuitive, but in the field of optimization, the bottommost part of that graph or that landscape is where we want to be. So maybe if you can analogize, you can think of the landscape as being cost or energy or distance. And the solution is how can I reduce or minimize the cost or the energy or the distance of whatever it is that I'm trying to solve. So if you can imagine, if I had a marble on this landscape and I wanted to go to the lowest point, sometimes the marble could get stuck in a divot, not even know that it's at the bottom and think that it's reached a minimum. That would be an example of what's called a local minima, not the global one. And that is an example of a case where the optimization did not quite work out. Another example could be if I'm a marble in a landscape and I hit a very wide flat plane and the marble doesn't go any further. Again, that's called a barren plateau. And it's an example where the system isn't allowing the minimum solution to be found. So one of the challenges that we have as any optimization algorithmist, especially in quantum, is how can we mitigate some of these challenges that are faced when we try to develop optimization algorithms. And we believe that this QITE algorithm is a step in the right direction toward avoiding these pitfalls. Next, please. On the next slide, we'll talk about one very specific plot from the paper that we recently published with Oak Ridge National Lab. You're welcome to read the paper by scanning the QR code on the right there. But really, what I wanted to talk about is really dig into this one chart to explain a little bit more what this optimization is about. So first of all, I'd like to circle back to a slide that Margaret showed earlier around our quantum computers. And our quantum computers comprise ions in a row. She showed that slide of ions in a row with 2 laser beams zooming on 2 ions. So in our quantum computers, we have something called operations, or gate operations, and that really involves the laser shining at 1 or 2 ions. So you can either have a single qubit gate operation where the laser shines at 1 ion or 2 qubit gate operation where the laser is split and shines on 2 ions and then those 2 ions interact with each other. So one of the real hallmarks, or one of the advantages around quantum computing is considered to be when those qubits are entangled with each other, so they are interacting with each other. That really leverages the inherent quantumness of the system and enables us to basically capitalize on the strength of quantum physics to help solve our problems. Another thing that's really important to consider is that as these lasers are addressing these qubits, you really want to try to minimize the number of operations. Why? There's really 2 reasons. The first is that the shorter the number of gates, the faster the time to solution. But also the shorter the circuit, the more accurate the solution. So really, we are motivated by the desire to get to the bottom of that landscape as quickly as possible and in the right well. So on the right-hand side, what we show in the dotted lines are actual examples where our algorithm was able to quickly get to the bottom of that landscape with fewer iterations than other methods. And this is, as I mentioned, one of the really exciting steps in the right direction toward coming up with a quantum optimization algorithm that can achieve these dual purposes of getting to the right minimum in the minimum amount of time. The other thing that I wanted to point out here on the right-hand side is we ran these on actual IonQ hardware, which is becoming increasingly important. I was looking at the chat here and one of the observers basically said that with the increase in the number of qubits, it becomes harder and harder to simulate a quantum computer using classical methods. We are rapidly reaching the point where classical computing can no longer simulate the number of qubits that we have. So when we get to that point, the option that we have before us is to be able to run this on hardware. And by showing this at this number of qubits, we are very confident that as our road map expands and we have more qubits, we will also be able to run these simulations on those QPUs as well. Next slide, please. Finally, recapping our pillar areas of chemistry, optimization, and AI/ML, I wanted to show a sort of buffet of the kind of hybrid algorithms that we've been able to develop in partnership with our customers. In particular, I've been here for about 2 years, a little bit more than 2 years, and I've seen some version of this slide or actually contributed to some version of this slide over the entire 2 years, and I'm really happy to say that this list is growing dramatically. And what I wanted to invite the audience members today is if any of these use cases resonate, or if you don't see your particular use case here, but would like to be part of this growth, please feel free to reach out to us, and we can see if we can do something together in this rapidly growing field. So next, I'd like to hand it off to Bjorn Flatt, who's the Head of Emerging Technologies, and he will talk about really the foundational layer of this full-stack, which is our hardware. So thanks, Bjorn.

Thank you, Masako. My name is Bjorn Flatt, and I lead the emerging technologies team here at IonQ. I'd like to take the next couple of minutes or next few minutes to talk about connected systems. In connected systems, we have actually several understandings for that. One, it's -- we just acquired Qubitekk, a quantum networking company. So connected system is clearly something we think about in the means of quantum communication. But it's also a way for us to scale our systems to connect our quantum, or QPUs, our ion traps to each other. And one step is shown on the next slide, is connect to come up with a modular architecture for our systems. In an earlier webinar in this past summer, we talked about our compact vacuum packages we're working on and that we were in the process of completing our UHV Assembly Chamber for prototype development. In this video on the right-hand side, you see that chamber, which has been completed in our facility in Seattle and is about to go into operation. This chamber allows us to produce thematically sealed trap packages. These packages will eliminate the need for mechanical pumps, drastically reducing the form factor and the volume of our vacuum systems. By removing mechanical parts, it will also increase the robustness and reliability of our IonQ systems, taking another step towards enterprise-grade capability. On the next slide, we see that we also designed our compact vacuum packages. In the graphic on the right-hand side, you see a comparison of the current package next to a quarter and compared to the first-generation vacuum package we used in earlier systems. So that's about a reduction in size of more than order of magnitude. This reduction in size, weight leads also to a reduction in power consumption. We do not have to power all these big vacuum pumps required there. Again, as a reminder, vacuum is important for us because our inherently pure qubits on the atomic base need to be isolated from environmental influences, and that's what we do in the trap package, in the vacuum package. Again, having these smaller, more compact packages allows us to accelerate the manufacturability, the installation, and the maintenance of our systems. Another advantage of these packages is shown on the next slide, where we point out that we can actually operate these packages at room temperature. And that is a huge reduction in the operational cost. If you compare that on the right-hand side, superconducting qubits, for example, operate at the tens of millikelvin range. If we can operate these at room temperature for our systems, we do not require cryostats, again, an additional reduction in cost, size, and weight. Also, we do not need coolants like liquid helium. That, again, is a step towards sustainability and manufacturability at scale. Overall, this approach for the compact modular systems together with interconnecting those packages, we have integrated photonics, which we addressed in a blog post about 2 months ago, are significant steps towards our 3-scale -- 3-pillar technology strategy for delivering performance, scalable enterprise-grade systems. With this, thank you very much, and back to Margaret.

Somebody has -- there we go. Some of you is watching TV, some very interesting television show. So this last slide, I wanted to just wrap up and talk about the summary of what we talked about today. At the core of what IonQ has been doing since the first of our building of our first computers, Harmony, and now Aria, which is our current quantum computer, Forte, which has been available, and AWS announced Forte Enterprise, Tempo, and future systems. And as Bjorn just talked about, we have to connect these systems because a really key part about scaling quantum as well as classical, to be honest, is connecting. A computer by itself is lonely. It's got no one to talk to, and it has nothing to -- it has no capability of getting bigger and bigger and stronger and more scalable without it being connected. There's also another notion of networking and connectivity in what we're doing with Qubitekk as far as getting a stack that relies on hardware, software and applications that then can actually be expanded and used in the more classical term of networking being able to go across a network. We did talk about our IonQ quantum software. It's an area that we hadn't talked that much about previously. But obviously, as with any computing system, we didn't need an operating system. We need the SDKs. We need the APIs. We need to be able to allow applications to be running on the stack, that is both the hardware and the software. Coleman talked in depth about hybrid services, very much like other areas when we used to talk about how do you get on-premises servers to actually communicate and work with other types of servers, let's say, on the cloud. We talked about hybrid, about being on-premise, hybrid in the cloud. In quantum, we have the same thing. How do we make sure that we are working with classical, working with companies like NVIDIA to make the ability to have quantum and classical work seamlessly together and be optimized to actually work so that any developer can easily use our stack. And then finally, on the very top, computers and software mean nothing if there aren't applications that have real outcomes and real impact, back to our mission statement. They're very complex problems that Masako talked about, complex problems that can save lives, that can change the world. We have to make sure that we have a robust applications team that's able to actually address some of the largest, most difficult problems with quantum algorithms that they're building. So that is IonQ's Full-Stack Quantum Innovation. We're very excited about the stack that we have been building over the years, and we have a lot more exciting things to come. So with that, I'll hand it back to Nikhil.

Thank you, Margaret, and thanks to our presenters. And thank you so much to our audience for joining today. We really appreciate it. At this time, we're going to take a couple of questions for the experts. So the first one is for Masako. Given the demonstrated success of QITE on 32 qubits for optimization, what advancements in hardware or algorithmic strategies will be required to extend its applicability to more complex problems such as those involving thousands of variables or constraints?

Yes, that's a really good question. And one of the actually phrases that was in the slides is that QITE is an algorithm that scales linearly. And that's a really important factor here. A lot of existing optimization algorithms, they scale polynomially, maybe even like high factor. And what that means is that if I had a certain number of qubits, the depth of the circuit scales more than linearly. It could be quadratic, it could be higher order. And what that means is that with conventional algorithms that we see for quantum optimization currently, the circuit becomes too deep. And as I mentioned earlier, when the circuit is deep, there's really 2 drawbacks. One is that it takes too long to get to the time to solution, but the other is that the quality of the solution degrades. So really, the promise of this particular algorithm is this linear scaling. And really, as the number of qubits increases, the depth only increases linearly. That does not mean that we're out of the woods yet. One of the things that Bjorn had mentioned is that in the future, as we have an increasing number of qubits, they're not going to be just lined in a row infinitely to 1,000, right? So currently, we have 35, 36 ions in a row. In the future, we'll have packages that are interconnected. And one of the great benefits of working for a full-stack company like ours where the hardware, the middleware, and the applications are all done in-house is that we can work with the hardware people to codevelop these algorithms based on our mutual need and mutual understanding. So that is something that we do very closely with our hardware team all the time, which is understanding and even specifying what the next-generation hardware should be based on the applications that we believe will be the most promising. So thank you for asking.

Thank you so much, Masako. This next and final question is for Coleman. How can I get access to IonQ Hybrid Services? And what immediate benefits will I see?

I love that question. I love that people are excited. So nearly all of the IonQ Quantum OS changes I talked about are currently running on IonQ Forte and its default production configuration. A few more we're still kind of finalizing testing for and will be rolled out over the next month. So if you have access to IonQ Forte, you already have all of this runtime acceleration available. You'll see speed. And I'll also note, like they, I think, just announced ahead of re:Invent, IonQ Forte is now available as an on-demand machine on Amazon Braket. So really, if you have an AWS account or a credit card, you have access to IonQ Forte. I will note, though, as a caveat that these speedups are workload-dependent. They're most dependent when running a hybrid job with air mitigation. Given that we've seen, like I said, at minimum kind of a 1.5x even despite the fact this is over a workload, similar workload run on an Aria system, which does not have these new capabilities because it is missing some of the firmware and control electronics components that it relies on despite Forte's actual gates being marginally slower. So for everything else, Sessions, solvers, functions, the SDK, we're not rolling those out in full general availability just yet, primarily just because as we're still iterating on some of these core workflows, interaction points, the programming model in its early days, we want to do that in really close collaboration with customers and partners. We think it's really important to get that really critical early feedback. And frankly, my team only has so much bandwidth. So we're limiting that pool of access. But anyone who's currently an IonQ partner or a customer, or wants to be, feel free to reach out. We can get you on that list. We still have, I think, some capacity to bring a few more beta partners on. Just reach out to your existing IonQ contacts or go to the link on screen to start building a relationship with IonQ.

All right. Thanks so much, Coleman. Thank you again to our presenters. Thanks so much to our audience. Again, a recording of the session will be available within the next few days. So stay tuned. And to get additional information about what you heard today or to learn more about how to get started, please visit ionq.com/get-started. You can also scan the QR code on the slide. You'll also get a brief survey after this. We really appreciate you filling it out, and it will take a couple of minutes. And thanks again for attending. We'll see you next time. Have a great rest of the day.