Daniel Rodan Legrain: Hello everyone and hello Laura. I'm Daniel Rodan Legrain, the Quantum Science and Engineering Lead for North America at Qblogs, the sponsor of today's Fisheye Chat. This series kind of brings together voices from across the quantum ecosystem to discuss not only where the technology is today, but also how we collectively build the future. In previous session, we've had the pleasure of hosting people such as Will Oliver, Berder John.
Laura Schulz: Hello.
Daniel Rodan Legrain: Peter Schorr, Irfan Siddiqui, David Afshalom, Lieven van der Sijpen, and many others, each bringing a different perspective on quantum hardware, algorithms, control, and the broader ecosystem. And today I'm particularly excited because our guest really sits on the intersection of several worlds that do not always speak the same language. mean, quantum computing, supercomputing, national labs, industry, and large-scale technology strategy. So Laura Schulz is the project lead for innovation at Argonne Leadership Computing Facility, where she focuses on future heterogeneous systems, integration and workflows. Before joining Argonne, Laura, you led the quantum computing and technologies at the Leibniz Superconducting Center, LRZ in Germany, as part of the Munich Quantum Valley. And there you really played a central role in quantum HPC integration. across Germany and Europe, including energy strategic plan for quantum computing and the Euro HPC joint undertaking project, EuroQ EXA. Even before that, you spent several years at Lawrence Livermore National Lab, where you worked at the intersection of high-performance computing, user facilities and industry-facing innovation. And Laura, you have really become one of the leading voices on this hybrid quantum classical computing. and the future of heterogeneous computing infrastructure. You are also recognized by HPCWire as one of the people to watch. So thank you so much for joining us today.
Laura Schulz: Well, thank you so much for having me. very nice introduction and thank you everybody who is listening in today and ver very, very happy to be with you.
Daniel Rodan Legrain: Thank you indeed. So for everyone watching live, please post your questions in the chat throughout the discussion. And I will save some time at the end as per usual to bring some questions from the audience into the conversation. So Laura, let's talk a little bit about your path. You've worked across national labs, superconducting centers, industry-facing innovation, and now really in quantum HPC integration. So how did your path lead you into this? intersection between classical high-performance computing and quantum computing.
Laura Schulz: that that's a great question. I think it was completely natural, honestly. so I was ahead of strategic development and and partnerships at LRZ before really getting into into quantum full on. And there it was always about emerging technology, paying attention to what's coming ahead, right? I mean as a as a it and and basically all HPC facilities and user facilities and service type institutions, you know, we're always thinking about what is on the horizon, what do we need to prepare for. So ideally when the users see a technology and say, that's that's, you know, I'm interested. I want to go into that. You know, as a as a provider, we don't want to say, okay, great, we'll get right on that. And then a few years down the road say, here you go. But we want to say, yes, of course, we've been tracking that, we've been following it. We realize that this may be an asset or resource that you want to use. For your science and your research. We're already working on this, you know, join us on this, or or here you go. Here's the resource. So always with that future vision, that future looking in mind, that's really where that came from. And so quantum, you know, we saw quantum coming down the pike. It was still in the physics laboratories, but there were very early startups that were branching out of the lab and into a startup environment. we definitely saw because this was a unique. an a a you know a really unique form of of calculation to to to retrieve some results in a very unique way, particularly to solve a particular portion of a problem, maybe part of an entire large scale science problem, we realized that we needed to start looking at this seriously, try to figure out what what was going on, how this could potentially fold into the ecosystem, into the advanced computing environment, what the problems would be, really just try to do an assessment and figure that out. So that's really how all that came about. And and, you know, the n the the noise level just kept getting higher and higher, you know, particularly in the in the European ecosystem where I was at the time. And so this really just became a part and parcel of of what an HP Center would would pay attention to.
Daniel Rodan Legrain: So maybe following a little bit on that, what's the specific moment when let's say you realize that quantum should not be treated as a standalone technology, as part of a broader advanced computing ecosystem?
Laura Schulz: Yeah. Yeah, that's a good question. I mean, you know, and and maybe it's from the the the time of you know that that I initially had at at Department of Energy where we think about large scale, you know, mission oriented, large scale science, right? And those science that science is those are science campaigns. These are problems that people have been working on for, you know, in some cases decades, right? And and constantly trying to get better understanding of you know, computational fluid dynamics, you know, that that relates to many different fields of, you know, biological problems, chemistry, physics problems. There are these, there's these large grand challenges, these large problems, and there's whole campaigns that were behind it. So, you know, we we've had modeling and simulation that has had this, you know, this this storied history and and evolution and it continues to evolve. When we look at AI, and we look at A new way of computing, a new way of having a different perspective on on a problem, taking part of that science problem and applying it to AI and getting a result that kind of gives you an overall an overall new solution or a new insight. And then we thought, okay, well, quantum, that's probably going to be the same way. We knew that there was going to be is going to be incredibly useful for particular parts of a problem. And so it just made, it just It just made sense that this was going to be part of the arsenal that we have for advanced computing and and how we were gonna bring that in. So I guess, you know, somewhat naively, I guess somewhat naively, I kind of came from the perspective where I I just didn't see it as as a standalone. I just saw the whole advanced computational ecosystem and okay, here's a great new thing that we wanna look at and new computing that's gonna be very different, very different. But what can how can we bring it into the fold? you know, and and that I mean, I was not alone in that space, obviously. Many, many people were in that direction, but that's that's kind of the origin.
Daniel Rodan Legrain: That's fascinating. Let me ask you maybe a little bit more of a personal question as you navigate through this transition. mean, you've moved through several different environments in the US, in Europe. What skills or experiences would you say ended up being most valuable that you did not necessarily expect early in your career?
Laura Schulz: Yeah, so I think so when I was at Livermore, you know, I I worked with I worked with that I guess that was my first exposure to thinking about large scale science, about mission driven science, where many, many, many people are are coming in and they're bringing their expertise and their experience and you know and they're they're working on a slice of the problem, right? And and so that that mission type science and space, that really stuck with me. you know come from from my exposure at Livermore. You know, when I went to when I went to Europe and into the Munich Quantum ecosystem, then that really became about taking these emerging technologies and taking all the lessons that we've learned, you know, from from, like I said, from decades of of building up advanced computing ecosystems and and how to pull in a new technology and really try to give it the the best the best scenario for success and you know really how to bring it in. I think that would be from what I gathered from from you know the the time of the Munich Quantum ecosystem. And then I think I would say from Argonx, I've been here a year now, it's w the technology is advancing and so now I'm thinking a lot about ecosystem design, about how all of this technology, how these different modalities are going to come together. So not just the HPC part, but the HPC with multiple quantum then also, I mean, as you know, there's a lot of focus on on AI with the Genesis mission, what that's going to do to, you know, to revamping or to to redesigning maybe some of our compute ecosystem. And so now it's this great time. So I think it's it's constantly a a synthesis, I guess would be the overall arching theme throughout synthesis and kind of different flavors that I've seen throughout my career.
Daniel Rodan Legrain: I mean, this journey really, you know, put you at the center of one of the most interesting questions in science and computing today. I mean, how quantum systems really fit into this broader, you know, computing landscape. So, you know, for many years, as I was training myself, would even say quantum computing was really discussed almost as a standalone technology. And you've been one of the strongest advocates for viewing it as part of the larger computing ecosystem as you've very well said. So how do you see this relationship between HPC, AI and quantum evolving, let's say over the next decade?
Laura Schulz: Yeah, okay. So you know, again, I mean I I'm I'm always thinking about the end user, right? And and I maybe that's from my experience from being in an HPC center, but we've always provided resources so that, you know, very intelligent people who are cracking really hard problems and trying to gain insight, they've got the tools and the and and and the and the resources to be able to do that. So I'm always thinking about, you know, the end user and and the and the the success ability for science, right? The output for science results. So, you know, when I think about how all of this is gonna play together, I mean, again, we've got, you know, HPC that's done modeling and simulation, that, you know, where we've been able to to simulate the natural world and and and and get some results that we can actually quantify and study. When we look at AI, we're we're amassing a lot of data, finding pattern and process in that to try to give us some insights. You know, quantum I see as being able to simulate results or being able to simulate quantum mechanical effects on an on a system that is a quantum mechanical device so you can, you know, simulate it there. And so instead of doing you know, we we we we've done this for for decades too, right? We've simulated quantum mechanical elements, you know, in HPC, but again, that's an abstraction. Right. And so now where that's really important, we can offload that particular part of a problem, put it on to the quantum environment, get you know more accurate, ideally I'm running understanding much more accurate results for that particular part of the problem, bring it back into the overall simulation, the overall problem, and and pr and hopefully gain some new insights. So I mean, I constantly see these playing on the strengths of each other. in into kind of an assembling into an entire whole. That's how I see it.
Daniel Rodan Legrain: Perfect. So let me dig a little bit more there. mean, when some people hear about quantum HPC integration, mean, they often think about maybe plugging a quantum computer into a supercomputer. But in practice, and maybe a little bit more technically speaking, what does integration actually mean?
Laura Schulz: Yeah, yeah, that that's a really great question. I mean I always I always talk about these layers because an integration, again, this is not just unique to quantum to HPC. That this is this this story just keeps repeating itself throughout, right? the integration means several different things. One, you know, you actually have the physical hardware. So regardless of what's happening with the quantum and the amazing job that the physicists are doing with the with the qubit fidelities and and the control electronics and all of that, right? You you have to bring it into an environment where you have to couple it with the comp with the classical compute, right? And so, you know, the early systems that that we've been dealing with, you know, have been physically planted in the HPC centers and we expect that, you know, to continue in in various gradients as we move forward. But that literally means that you're taking what was a physics experiment, you know, lovingly cared for by you know PhD grad students and and technicians and people with physics degrees, bringing it into an ecosystem where we don't necessarily have those expertise and resources. But it's also the sensitivity of the system and the sensitivity of the system in an HPC environment where you've got, you know, very large systems running, different different facility infrastructures, all of that sort of thing. So you actually have to look about how you bring those systems in. And that and that includes everything from you know shielding it, and and and again, the the what the companies are doing to actually shield the systems, but then what the facilities' role is in in maintaining, making sure that those systems operate correctly. And that You know, that that's something that the companies can't completely do on their own. They do they do need the understanding of what the facility infrastructure and environment looks like. So you've got that point. you actually have like the connec the actual connectivity of the systems, right? So the the what sort of networking, the the actual architecture, where are you going to put like the common data? How are you gonna do the networking together? What is the what how are these gonna systems gonna communicate together? You've got the system software environment. That's kind of like the next layer up. And you know, this is non-trivial because now you've got quantum, which in HPC land we would call a disaggregated accelerator because it, you know, it has its own operational environment and it sits here. it's not fluid like what we can do with with CPUs and GPUs. it's a disaggregated accelerator, so you've got to deal with co-scheduling, and you typically tend to have you know, several different schedulers in in in accelerators like this. So you've got to coordinate that. You have to res manage the resource. You have to figure out the timing of when the job is on the HPC side and then parts of it go over to the quantum side. What happens to the to the jaw or to the the nodes that are on the HPC portion, what w you know the time that it takes for them to come back. You ha have to make sure that you're not keeping systems idle or, you know, or just waiting around for jobs. So how do you manage that? There's that whole level of of system software. That's even before you get to the application layer and before you get to the user culture. So there's all these different layers that you have to think about here and how all this works together. And then the other component, which is kind of a new component because we actually have systems launching now, is that now you've got systems distributed all across an ecosystem, whether that be regional or national or even international. What is going to be the coordination between those those quantum systems, between the quantum system and the HPC system that might be co-located or not? How does all that work together? So this is a really it's it's fascinating because there's several layers here and there's a whole bunch of questions that we're still actively researching at all of those layers. And Some of them depend on, you know, one layer, the answer is dependent on another layer's answer. And so you kinda have to to really think about how all these pieces come together. So integration is not just plugging in a system and, you know, plugging an Ethernet cable to it and and waiting for it to go. There there's a lot involved here.
Daniel Rodan Legrain: I can see a lot of open questions. So you mentioned all these layers. What would you say are the first layers that need to be integrated? you say facilities, scheduling, software, workflows, user access, something else?
Laura Schulz: Yeah, I mean that's that's a great question. And it's I hate I hate to be one of those, well, everything, yes, but but in some ways that that is true. I mean so like you know, early on, yes, we we had to get an understanding of the of you know, actually if a system was placed in in a system or in an ava HPC center, what sort of questions do we have? that's one component. But but already I I guess kind of answering this sorry very vaguely Every single layer has, you know, kind of critical steps that we have to take. And so to me it's kind of mixed. And and I think that that's actually important too because we're it's gonna be informing the other layers. And this is really when we talk about co-design, which is a word that, you know, that in in HPC space is is kind of our bedrock to to how we operate. When we look at the applications, when the applications team are thinking about how to put these systems together, how to how to utilize the compute from both of them, if it makes sense for their for their workflow for their application, they're gonna have particular sticking points. They're gonna identify bottlenecks, right? that's gonna influence what happens, like how you design the system software environment. you don't you you don't want you wanna make sure that you're developing for the user, you're not the developers aren't driving in and just telling the users here. figure out how to use it, right? And so there's this con as all of these lines and all of these people involved are are are working in these different spaces, they're going to end up hitting different parts of those layers. And so all of this is gonna work together. So it's a ri really long winded saying to way to say it all comes together at the same time, I think.
Daniel Rodan Legrain: And where are we today? mean, is it proof of concept integration? Are we in early production systems or are we somewhere in between?
Laura Schulz: So I we are we are in proof of concept, but a lot smarter than we were when we were doing proof of concept a year or two or three ago. so systems are actually, you know, coming into centers. And and this is not, I mean, this is for the the I mean I'm I'm really focusing now on this the quantum systems that are in the HPC centers. I mean, we you know, on the application layer and all, we've been doing a lot of discovery and insight. remotely accessing systems, you know, working on those, working with those vendors, doing all of that as the technology is developing. But I say would I would say we were proof of concept, but smarter than we were before. We're asking better questions. And you see that from the conferences. I mean, I was just at ISC a couple of weeks ago and I was honestly a bit stunned at at how much it was quantum this time. you know Four or five years ago when we held some of the first workshops in this space, to be really honest, we were kind of saying, Yeah, we're we're gonna need to, you know, figure out how to do this. And we were all honestly kind of slow blinking each other in in the sessions and and some of the papers that we had were it was very rudimentary. Every single conference, ISC, IEEE Quantum Week, SC, you know, you just constantly see more quantum and deeper. Deeper questions, deeper research, more people coming into the space. So I I would say that we are well on our way into the proof of concept, early, early deployment and all of the information and knowledge that that brings to us.
Daniel Rodan Legrain: Amazing. Well, I look forward to all these conferences coming our way. So Laura, from your perspective, what would you say have been the biggest surprises in trying to make that integration real throughout all these years?
Laura Schulz: I don't say I I wouldn't say it's a I I guess I wouldn't say it was a surprise, but it just reminded me of how hard we have to work. And and that is on the the translation and the and the communication. Right? So with with any project in general, I and project managers know this that basic I mean, there's a statistic from you know from your Pembok books that that like ninety percent of any project is is communication. You know, if you're you know, trying to to to coordinate and and develop things. And that's really true because we have such different communities that have you know kind of the starting point, with different cultures, different perspectives, different operation modes, di you know, different behaviors, bringing them together, but at the same time everybody is evolving and everybody's growing. So the, you know, the the quantum hardware companies They're actively working on their stuff, the the on their technology. The the software is trying to you know build this environment, the users are trying to figure out what to do. There's all these pieces, and because of the co-design part, the communication is paramount. And the translation of these different cultures and different needs and different user bases. And so translation, communication, coordination. that's always important. But because we have such a dynamic, fast-paced environment right now, that's even more important. so keeping everybody on the same page or at least with situational awareness of what's going on, and and and hopefully not letting factions silo too much into, you know, into their own solutions, right? That that then we have a hard time bringing together, you know, these technologies for the benefit of the user and their science. So I would I yeah, translation, communication, coordination, way more way more effort than than even I think a lot of us thought.
Daniel Rodan Legrain: I yeah, I guess that's a massive project that will really require a strong community coming together. So along these lines, Lara, I mean, you've mentioned all these layers that need to come together, but what do you see, let's say today, the biggest bottlenecks preventing this seamless quantum classical integration?
Laura Schulz: well for one, I mean, well, a few things. I mean, one, just the deployment of the systems, right? So we're we're s we're slowly having systems be deployed, but honestly, worldwide there's not that many that are connected, quantum systems connected with a through a s a a concrete through path with a system software environment to to an HPC system. So that we can do some of this work, right? And that and that's gonna come more as as time moves on. but that's part of it. You know, there's only certain places where this work, you know, can really be done, these hot spots. there's that part. And then of course, you know, in the the the big thing that the quantum technology itself is going to be going through a paradigm shift here very soon with you know with the emergence of fault tolerant computing. what that's gonna mean, you know, for the for the connection, the integration with with classical compute. There's the the application workflow stu flow there. Then there's also the error correction classical compute support of the quantum system and how those are going to how those are going to manifest within the same or different computing resources. but then but then you know How how are we going how are we going to manage the the various classical compute systems? So I think the rapid evolving of the the system at the quantum technology itself, but then also the the elephant in the room that's you know that's breaking the door down here soon with fault tolerance. And then what's gonna happen? With what we've already done. So NISC era was both good and bad, right? So there was a lot that we learned, but there's a lot that we're probably gonna have to unlearn, and that's just part and parcel. But then what can we reutilize? What can we learn from NISC and apply to fault tolerant in terms of again, the whole layers, the system software environment? Is can we take a whole lot? Maybe, hopefully. but in terms of application. So all of these layers coming up, we're gonna have a shift as we go to to fault tolerance.
Daniel Rodan Legrain: So maybe a little bit of a controversial one here, but do you think quantum systems will need to become more HPC-like or do you HPC centers need to become more quantum aware?
Laura Schulz: gosh. I I hate I hate saying both. I I really don't want to hedge my bets, but but but it is really true. you know, we when we when we think about, I mean, HPC supercomputers and that and that environment, these are very bespoke, well tuned systems, right? And they already have their own Operational demands, and I mean it's not like you just plug these things in and they work, right? I mean, there it's it's it's very mature technology, but it's continuously evolving itself, and so these are very balanced systems. So you can't just say, you know, quantum, you know, you guys just need to put on this sort of plug in for this, you know, this new plug in for this scheduler, or do this, or use a different schedule, or whatever. You can't have a unique solution. That works with your quantum system and then toss it over the fence to the HPC part, right? That that doesn't work. This all has to work really part in parcel. And so and so the quantum environment has to really understand, you know, what it takes to integrate to the HPC side, what those previous patterns and processes are that you know that it needs to abide by. It doesn't have to abide by all of them. You know, we we want we want some flexibility and innovation. But there are some guardrails that we need to respect. You know, on the other side, this quantum is just a completely new way of approaching computational results from a system. And we have to honor the physics of these systems and what they're going to require that's going to be delta to what we've what we've dealt with in the past. So there's gonna be learning and accommodation on both sides for all of this to work.
Daniel Rodan Legrain: Very fair, Laura. Let me switch gears a little bit. for most users, would say that the experience of a system is often defined less by the hardware itself and more by the tools, the abstractions, and the workflows around it. So let's talk about user experience. From your perspective, what are the biggest barriers to adopting these hybrid quantum HPC workflows today?
Laura Schulz: Yeah. Well, I mean, it's again the the just having the the actual ecosystem set up, you know, that that's something that we have to really work on. but then, you know, again, as I kinda mentioned before, you know, we're coming from from two different cultures. So one, you know, the HPC folks who have been doing this work and and and doing simulations of of they're gonna want to understand what what can I actually do with quantum. you know what what is the what is the actual behavior of the c n not even so much I don't want they don't want to know intimate details of the qubits right they want to understand you know behavior wise this you know when I decompose my problem this is what's happening at this point in the steps where can I push this off to quantum when does it make sense what's the benefit how is it actually beneficial over what I'm already doing Right. And so they there there's gonna have to be a lot of awareness building, decomposing the problems, understanding where the quantum fix fits in. That's that whole part there. So they're gonna have to get a lot smarter on the on understanding what the compute can actually do for them. on the on the quantum side of things, you know, to really understand, you know, where the classical can help with the pre and the post processing, where once once you've had the the power of the quantum compute and gotten what you can get out of you know out of the the run with that how can you bring it in bring it back to your overall problem how can that system support you for for your quantum for your quantum calculations so it's it's it's funny at at ISC we had a a a boff on on workforce development and what was interesting to me I asked the question to this group and I said you know the physicist raise your hands what do you want to know most, you know, moving forward about like integration? And all of them said in some flavor or another, more understanding of what it is HPC is and and you know, what it means and and and how I'm supposed to think about it and what I can do. And then reversely, the HPC post were like, Yeah, we need to understand more of the quantum side of things. And right now we're at the point that they actually have to get get get cozy with the qubits and they need to understand kind of hardware at a level that we wouldn't ask them to understand if they were in the HPC space. You know, right right now, we have so many abstractions built up in the HPC realm that we try to we remove the users from intimate understanding of of the bits and the transistors and all of that, right? But on the other hand, right now, because of where we are with quantum, people who use quantum kind of need to understand this and have a you know have an understanding of how to operate it. So what we want to try to do is remove that away from them as much as possible and build an environment that they understand they understand enough of how the compute works and behaves, but we don't want them to have to be quantum experts in this space in order to be able to use it. That's these folks are experts, but that that's that's quite a big ask and that's that's quite a big bridge to cross.
Daniel Rodan Legrain: Indeed, yes. And how do we give users, let's say, a common experience across different quantum technologies altogether? I superconducting qubits, neutral atoms, trot ions, spin qubits, and others.
Laura Schulz: Again, a fantastic question. I mean, I think the first thing is literally just the access and the you know the democratization of the access. So hands-on is the most powerful thing that we can do. and and already we're already starting to talk about infrastructure. how how do we how do we provide access for these systems on a common back plane that also has, you know, at the front end it has the same software environment, right? And that they can utilize. regardless of the technology that's common enough, you know, to get them to a certain level, but then it also honors the unique physics of the individual systems when that part becomes important, if if that makes sense. So it's you know, it it's about making sure that we we get the access, but that we're providing the ecosystem that they can use it. I mean, what what's happening, and this is this is completely normal for for early days of technology. We saw this in HPC. We we, you know, this is a reoccurring pattern where you you know have technologies that come up and and everybody wants to build a full stack solution, right? From the hardware to how it operates to the application software that it uses. And but what you end up getting is a whole bunch of different users on one side, a whole bunch of different devices on the other, and then you get this Hornets nest of workflows and tools and all that aren't really all working together. And the problem that that causes is that you're gonna end up getting a silo effect or you're gonna end up getting you end up getting necessary allegiance to one technology and one, you know, one s series of resources and a workflow because that's where there's been the investment, either in your center or you know, from your chair or whatever it is, right? So you end up kind of get locked in to particular areas and maybe that's not the best for your Applications for your work. We don't want to do that. So kind of the role that we have as HPC centers and you know and also would say as governments is to figure out a way to have to provide access to these different resources, but to provide kind of as much of a common middle ground with that software environment, you know, coupled with the application side so that these users can experiment. And be able to understand the best technology for their particular problems.
Daniel Rodan Legrain: Fantastic. Let me switch gears a little bit, again, and let's talk about ecosystems and why they matter. mean, you've worked deeply in both the US and the European advanced computing ecosystems. So how would you compare their approaches to quantum HPC integration?
Laura Schulz: that's really a great question. lots of great questions, Danny. on so on the on the European side, you know, it it started off with and because of the you know the really impressive efforts of like Euro HPC started off early with providing quantum hardware startup companies, you know, with with first first system acquisitions. Putting them into HPC centers, really with the mandate early on. Here's a quantum system. Figure out the technology is still rapidly evolving, but go ahead and start figuring out how this is gonna be c part c become part of the computing ecosystem. Right? And and and so yes, it's still evolving, but go ahead and start connecting it. figure out figure out what you could do. So this really this really facilities level, ecosystem level, hands-on, you know, direction. the the US side is and the reason that I you know love DOE so much is is because of this big mission directed science, right? And so really looking at of the applications and things that we that we have a priority To work on where does quantum come in? So I think it's coming heavily from I think it's coming from the the the HPC center side. I would say the US side was coming heavily from the application side. What can we do with this? And really trying to experiment under the hood and testing, you know, virtual systems first or cloud-based systems. what we could do on the application side, but then And then on the European side it was very hands-on with the technology, get it into a center and start building from there. So I don't want to say top down and bottom up, that's that's not correct. But it was kinda but it I do kind of feel like there was two ways of looking at it, and I think they're fully complementary and and I and I do think that it works very nicely together.
Daniel Rodan Legrain: Fascinating. So how can the US and Europe, let's say, learn from each other as they build quantum H-15 infrastructure?
Laura Schulz: Yeah, I mean I again that they're gonna come from they're gonna come from different perspectives, but at the same time we're also, as we kinda mentioned before, I think the the beauty of this of the of quantum HP C of this of this little hybrid environment that we're building up is that we all are working together really well right now. I I really enjoy this space because I think it is very collaborative and and cohesive. so I think that the you know what What the US side and these big, you know, the big companies that are really pushing forward and have some great results, the the the focus on applications, you know, what they're gonna be bringing to this space and what the the European and also we can't forget the the Japanese space, which is is is really amazing and and Asia in general is really is really kicking up in this space. Everybody's coming at this from a slightly different starting point, a different vantage point, and and maybe a different strategy. And so Bringing those together, yeah, th this this is really powerful and and and very useful. and we're going to learn from each other and you know we're gonna do things differently, and that's okay. What we wanna try to do is we want to try to we don't you know we talk about standards too much. We don't want to talk standards right now, it's all still too new, but we wanna talk interfaces. we wanna talk the ability of where are there particular couple, you know, coupling points, where can you connect in To this overall environment that we're building up with the software, with the hardware design or ecosystem design, connectivity, all this. Where can we find certain points that it's easier if you just kind of follow this particular interface or contract, be a simple API or a more advanced, you know, understanding of that? Where can you kind of connect into this so that we can have multiple different approaches, but still not silo? Right. I think that's gonna be really important. And and gl I think quantum is just very global. I mean, I mean, y you know this. I mean, QBlox you know, is in Europe, but it's in the US. It's working across like the entire global ecosystem. Pretty much all companies do this in some form or another. So I think everybody's a part of these multiple ecosystems. And so I and that that's also very encouraging that this is all going to eventually work together very nicely.
Daniel Rodan Legrain: amazing. And let's talk about the people because none of this works without the people. mean, throughout your career, you've really built all kind of multidisciplinary teams spanning from what I heard, physicists, computer scientists, software engineers, mathematicians, system architects, you know, all kind of user facing teams. And I'm sure many other people coming together. So what have you learned about, you know, leading teams where, you know, people often speak very different technical languages? especially in such a complex space.
Laura Schulz: Yeah. Yeah. Yeah. So I mean the thing is is like we don't want to we don't want to generalize everybody out, right? You so you need you still need those experts in each and every space. the the team that I was privileged to lead before was was exactly that. It was this group of experimental and theoretical physicists, computer scientists, software engineers, you know, y developers, electrical engineers, you name it. And everybody's expertise, you know, came into that in c into that pool. You know, I mean we it's not a melting pot, I think it's a salad, right? And so, you know, these people have all their different unique backgrounds. But then again, it it comes down to spending a lot of time talking to each other, understanding each other's perspectives. And you know, and we see this I mean this wasn't just like my the like an individual team at a at an HPC center. There's an initiative led by Amir Shahada in at Oak Ridge called OpenQSE, right? And that and these, you know, working like bringing all of us together in that ecosystem. And again, the same sort of thing, you know, from the hardware companies, you know, software developers, et cetera, control electronics companies, really trying to sit down and and build common interfaces, common, you know, points where we can all agree that you know our technology should do this at this point and and that we should be able to work together. So it's I and I also think there's gonna be a really interesting really interesting for folks in the future when we talk about workforce development because we still need those experts. You know I mean you still need people that we can go to to say this person knows about you know about fidelities with the qubits or or knows what happens, you know, for the with its error correction. At the same time I think everybody is going to have to have the ability to zoom out of their prospective expertise and have an understanding of what's going on, you know, in other spaces. And and not this is not just applied to HPC and quantum. I mean, I can tell you, sitting in on on discussions with with AI, you know, and and building up the AI ecosystem, there's a lot of work. that's already being done between HPC and AI integration and what that means. And and there's already big investments happening in there. And so there's already processes being built up to solve that particular, you know, connection problem or or whatever those issues are. And we need to be able as a quantum community to understand what's going on there, leverage what we can again for this whole ecosystem You know, again, for the benefit of science, we want to have end-to-end workflows. So we need to be paying attention what's happening there. And then at the same time, we need, you know, the AI and the HPC people to be actively in this conversation. So I guess one of the biggest things and that I always say is when you go to a conference that's like SC or ISC, you know, go to the quantum things for sure, but go spend some time over in some of those, you know, pure AI and pure HPC. Because I I guarantee you you're you're going to hear things that resonate with what you're with what our community is working on for sure.
Daniel Rodan Legrain: Duly noted, Laura. I see quite a lot of questions popping up in the chat. So let me maybe end this conversation from my end with a broader future looking question, and then we can switch to the audience questions. You describe the future of computing as increasingly heterogeneous. So if we sit here again 10 years from now, what does state of the art computing look like by then?
Laura Schulz: if I had my wish, first of all, I mean, fault tolerant is is you know gonna be, you know, moving forward, we're going to have those first systems. I mean, for me, we would have more systems with HPC connected to it. we would have a a deeper understanding, we would have a more evolved and matured software environment. But kind of all of it comes down to there would be some really interesting use cases. There would be some really interesting applications that people have sorted out that they did by asking questions they couldn't have asked before because you know the environment wasn't set up for them completely. So you know we we had some early, some really just interesting, early, very early, understandably, you know, results from when we hooked up the first quantum computer to the Supercomputing Center in Munich. Where I worked, right? The system. And rudimentary, rudimentary software, completely understandable. However, it already changed the way that the questions were being asked and ideas were being formulated and thoughts about the workflow. So the more technology we set up, the more infrastructure or ecosystem that we set up, the more mature it comes, the more interesting questions and results that are going to happen. So I just want to see really great science.
Daniel Rodan Legrain: Fantastic. Let's look over to a great science. Thank you very much, Laura. That was a really fascinating conversation. Let me bring in a few questions from the audience in the last few minutes. I see quite a lot of questions. Let me start by timeline one. So will hybrid quantum classical cloud services become more common in 2026, meaning now to 2030? or is it still a little bit too optimistic?
Laura Schulz: Really good question. so I mean w when when we say cloud, we just mean a facility that would have access to a quantum system and access to an HPC system and have the capability to put them together. I mean, I think in terms of very rudimentary integration, that's I think that that's achievable, right? when we talk about more nuanced and more mature and complex integration. That's still gonna take a little bit of a while for us to do. One, because we're still learning how to do it. But but yeah, I don't think it would be I don't think it would be out of order by any means to start seeing you know, some sort of access to rudimentary HPC and cloud compute on on on cloud services.
Daniel Rodan Legrain: Thank you very much. Here's another one. You've mentioned it's important for quantum and HPC communities to really understand each other's technologies. What are some good resources for quantum physicists to better understand about HPC? And if I may add a second part of the question here, what do you think new generations of scientists that want to enter the space should be learning and focusing on?
Laura Schulz: This is just gonna be a contact sport, to be really honest. So I think you know, there's been a lot of educational resources built up for kind of quantum-only understanding, you know, so how to do this. What what I do think that we need to see more of coming up is quantum, quantum explained for HPC practitioners, right? And and kind of taking N let's not start completely with the concept of you know of of qubit and superposition and all these. Let's let's really look at it, you know, when you're when you're dealing with you know with your code and you're dealing from your C or your C environment and and and how you're thinking about you know your your your code decomposition and and everything that's happening. This is this is literally the the computational behavior that quantum offers. This is how it operates, this is what you can do. So there needs to be how I say it this right way. There needs to be HPC coded language on quantum, right? I think that's kind kind of the big part. that needs to happen. And and and I'm hoping that that we start having more educational courses from that direction, taught from the HPC perspective, what quantum means, not so much from the quantum community trying to explain, you know, to the HPC. Does that make sense? it's coming from that aside. I think that's gonna be really important. So I'm I'm hoping that that a lot of folks kind of pick up that mantle and and and drive on that pretty hard. on on the quantum side, I mean the the the the kicker is you know in the physics community there are a lot of people that are that are that ha that are using HPC, right? So I mean, you know, you've got the quantum chemists and and you've got these folks who they've been utilizing simulation on HPC systems to understand, you know, some of their quantum mechanical effects. So they already so there's a lot of folks that already have that knowledge, right? And so now they're basically being able to take it, take part of that problem onto a quantum system. There's that class. For the pure, pure physicist that that haven't done the HPC or haven't really done the that sort of computational simulation part, yeah, that's a great question. I I wanna say very blithely that it almost seems like you have to just start with HPC, but yeah, I might have to think about that. Like how how do we onboard the the the physics folks for the HPC that maybe don't have the understanding of HP C simulation? I'll I'll have to get back to you on that one.
Daniel Rodan Legrain: Sounds great. There's actually a similar question, but in the other direction. How do we get people from the traditional HPC industry to really understand which algorithms are, let's say, most effectively suited for quantum?
Laura Schulz: Yeah, I mean this this again, and and I mean maybe I'm coming at this kind of naively, but I mean g once you start putting quantum on anything, you know, there's there's this like, my goodness, you know, this is gonna be so hard. And part of that is because again, it's been wrapped up in the in the culture you know, of very talented physicists who understand, you know, who understand the lingo, who understand all these components, and there's a translation problem to the to the HPC folks. So to me, it always comes down to behaviors. What I would like to understand is like, for example, in in HPC space, I mean we we do have HPC algorithms, right? And we have HPC algorithms that can that have perform a particular function that can be, I don't want to say reply replaced, but that we can we can take HPC versions of the work that the algorithms were doing there and and utilize quantum algorithms in their place, right? So I guess one thing would be, let me put it this way, I think probably the best thing to do is to get the HPC algorithm developers and the quantum algorithm to get developers, put them in in a really nice location for a long period of time and lock the doors. because I th I think I think that there's gonna be a need to be a lot of sitting down and figuring out at that level to be able to understand, you know, to to understand what the potential would be.
Daniel Rodan Legrain: I like the idea.
Laura Schulz: But but but it'll be really good it'll be really good pizza, you know, the food and it'll be really nice. But I again this just comes with the idea that it just has to we have to we have to we really have to put these communities together. E we're doing a great job, folks. Everyone's doing a really nice job of all of this. We have to do it even more so as we move forward.
Daniel Rodan Legrain: Fantastic. Let's do it. Let's talk to the people that have the power to make it happen. And I think you're sitting among them and you're part of them. Laura, I think we have time for one more question. So do HPC professionals and researchers like to collaborate with full stack quantum computing companies or any other companies in the value chain like software, QP, developers, control electronics? From the HPC point of view, who is the easiest or more practical to work with?
Laura Schulz: so Well, I mean I I think I I I'll give a shout out to integrators in this space now. So you know, when like again, HPC systems are are very complex entities, right? I mean these these supercomputers are an acc an an accumulation of of you know various network technologies, of of data store technologies. of the not just the you know the the CPUs the GPUs the telemetry environments the the the power part all these are all pieces that come together to build a supercomputer right and and companies who do this sort of thing are are integrators who can pull all these different vendors together all these different solutions work with you know the HPC centers to figure out what the best solution is and then and then work with them to map out what that solution is going to look like and then to and then to actually operate it. So I do see a big role for integrators with quantum companies because they're going to be the ones that are also going to look at the various technologies, the various solutions, again trying to you know work with the the the clients at the end what is the solution that's needed and then how are those components gonna fit in. So I mean I think integrators are a are a big one. And then there's the there's like the hardware integrators and the solution integrators, but then there's also, you know, software integrators. you know, so I mean again like one of the examples, not not a shock, would be, you know, something like NVIDIA and some of the stuff that they're doing with quant because they understand the software environment from the quantum and the HPC. Right. So it's again anybody who is dealing with Thinking about all of this in whole and not as individual, that's gonna be the sort of companies. And again, there's also quantum companies that are thinking that way as well, right? That are moving over. So I guess it it's interesting, it's not really a one or the other, it's more a philosophy and approach to how the company is thinking about this. Just buying a generic or just buying a quantum system that just works as a quantum system. that puts a lot of burden on the on the maintainers and the operators who are purchasing it, right? So those who are actually working on the hybrid solution are are the ones that we want to that we want to encourage.
Daniel Rodan Legrain: Fantastic. We definitely need more integrators, Laura. Laura, thank you so, so much for joining us today. mean, your perspective is really valuable. You've helped really shape this field from all kinds of different angles, national labs, superconducting centers, European initiatives, quantum HPC integration as a whole, user facilities, industry collaborations, and all kinds of international programs. One of the themes that really stood out throughout this conversation is that the future of computing will not be built through one single technology alone, and there's many layers to think through. It will be definitely be built by bringing together all kinds of different technologies, but also different disciplines and different communities, quantum, HPC, AI, hardware, software, facilities, users, you name it, many others.
Laura Schulz: Yep. Yep.
Daniel Rodan Legrain: That, I would say, connects very naturally to what we often say at QBlocks, building quantum together. It's not just a motto. It's a recognition that useful quantum systems will require a whole ecosystem. And I think today's conversation has been a fantastic reminder of that and how it looked like in practice. As a small token of appreciation, Laura, we've sent you a QBlocks mug with our motto, building quantum together. Hopefully, it accompanies you on the next round of strategic planning at Argonne. To everyone that joined us today, thank you very much for spending the hour with us. If you join later or want to revisit the discussion, the recording will be available on the QBlocks YouTube channel and the audio version will also be available on Spotify and Apple Podcasts as part of the Quantum Builder series. Stay tuned for more exciting Fireside chats coming soon. And Laura, thank you again and thank you everyone for joining us today.
Laura Schulz: Thank you so much for having me, Donnie, and the Q Blocks team. Thank you.
Daniel Rodan Legrain: Our pleasure, Laura.
Laura Schulz: Hello.
Daniel Rodan Legrain: Peter Schorr, Irfan Siddiqui, David Afshalom, Lieven van der Sijpen, and many others, each bringing a different perspective on quantum hardware, algorithms, control, and the broader ecosystem. And today I'm particularly excited because our guest really sits on the intersection of several worlds that do not always speak the same language. mean, quantum computing, supercomputing, national labs, industry, and large-scale technology strategy. So Laura Schulz is the project lead for innovation at Argonne Leadership Computing Facility, where she focuses on future heterogeneous systems, integration and workflows. Before joining Argonne, Laura, you led the quantum computing and technologies at the Leibniz Superconducting Center, LRZ in Germany, as part of the Munich Quantum Valley. And there you really played a central role in quantum HPC integration. across Germany and Europe, including energy strategic plan for quantum computing and the Euro HPC joint undertaking project, EuroQ EXA. Even before that, you spent several years at Lawrence Livermore National Lab, where you worked at the intersection of high-performance computing, user facilities and industry-facing innovation. And Laura, you have really become one of the leading voices on this hybrid quantum classical computing. and the future of heterogeneous computing infrastructure. You are also recognized by HPCWire as one of the people to watch. So thank you so much for joining us today.
Laura Schulz: Well, thank you so much for having me. very nice introduction and thank you everybody who is listening in today and ver very, very happy to be with you.
Daniel Rodan Legrain: Thank you indeed. So for everyone watching live, please post your questions in the chat throughout the discussion. And I will save some time at the end as per usual to bring some questions from the audience into the conversation. So Laura, let's talk a little bit about your path. You've worked across national labs, superconducting centers, industry-facing innovation, and now really in quantum HPC integration. So how did your path lead you into this? intersection between classical high-performance computing and quantum computing.
Laura Schulz: that that's a great question. I think it was completely natural, honestly. so I was ahead of strategic development and and partnerships at LRZ before really getting into into quantum full on. And there it was always about emerging technology, paying attention to what's coming ahead, right? I mean as a as a it and and basically all HPC facilities and user facilities and service type institutions, you know, we're always thinking about what is on the horizon, what do we need to prepare for. So ideally when the users see a technology and say, that's that's, you know, I'm interested. I want to go into that. You know, as a as a provider, we don't want to say, okay, great, we'll get right on that. And then a few years down the road say, here you go. But we want to say, yes, of course, we've been tracking that, we've been following it. We realize that this may be an asset or resource that you want to use. For your science and your research. We're already working on this, you know, join us on this, or or here you go. Here's the resource. So always with that future vision, that future looking in mind, that's really where that came from. And so quantum, you know, we saw quantum coming down the pike. It was still in the physics laboratories, but there were very early startups that were branching out of the lab and into a startup environment. we definitely saw because this was a unique. an a a you know a really unique form of of calculation to to to retrieve some results in a very unique way, particularly to solve a particular portion of a problem, maybe part of an entire large scale science problem, we realized that we needed to start looking at this seriously, try to figure out what what was going on, how this could potentially fold into the ecosystem, into the advanced computing environment, what the problems would be, really just try to do an assessment and figure that out. So that's really how all that came about. And and, you know, the n the the noise level just kept getting higher and higher, you know, particularly in the in the European ecosystem where I was at the time. And so this really just became a part and parcel of of what an HP Center would would pay attention to.
Daniel Rodan Legrain: So maybe following a little bit on that, what's the specific moment when let's say you realize that quantum should not be treated as a standalone technology, as part of a broader advanced computing ecosystem?
Laura Schulz: Yeah. Yeah, that's a good question. I mean, you know, and and maybe it's from the the the time of you know that that I initially had at at Department of Energy where we think about large scale, you know, mission oriented, large scale science, right? And those science that science is those are science campaigns. These are problems that people have been working on for, you know, in some cases decades, right? And and constantly trying to get better understanding of you know, computational fluid dynamics, you know, that that relates to many different fields of, you know, biological problems, chemistry, physics problems. There are these, there's these large grand challenges, these large problems, and there's whole campaigns that were behind it. So, you know, we we've had modeling and simulation that has had this, you know, this this storied history and and evolution and it continues to evolve. When we look at AI, and we look at A new way of computing, a new way of having a different perspective on on a problem, taking part of that science problem and applying it to AI and getting a result that kind of gives you an overall an overall new solution or a new insight. And then we thought, okay, well, quantum, that's probably going to be the same way. We knew that there was going to be is going to be incredibly useful for particular parts of a problem. And so it just made, it just It just made sense that this was going to be part of the arsenal that we have for advanced computing and and how we were gonna bring that in. So I guess, you know, somewhat naively, I guess somewhat naively, I kind of came from the perspective where I I just didn't see it as as a standalone. I just saw the whole advanced computational ecosystem and okay, here's a great new thing that we wanna look at and new computing that's gonna be very different, very different. But what can how can we bring it into the fold? you know, and and that I mean, I was not alone in that space, obviously. Many, many people were in that direction, but that's that's kind of the origin.
Daniel Rodan Legrain: That's fascinating. Let me ask you maybe a little bit more of a personal question as you navigate through this transition. mean, you've moved through several different environments in the US, in Europe. What skills or experiences would you say ended up being most valuable that you did not necessarily expect early in your career?
Laura Schulz: Yeah, so I think so when I was at Livermore, you know, I I worked with I worked with that I guess that was my first exposure to thinking about large scale science, about mission driven science, where many, many, many people are are coming in and they're bringing their expertise and their experience and you know and they're they're working on a slice of the problem, right? And and so that that mission type science and space, that really stuck with me. you know come from from my exposure at Livermore. You know, when I went to when I went to Europe and into the Munich Quantum ecosystem, then that really became about taking these emerging technologies and taking all the lessons that we've learned, you know, from from, like I said, from decades of of building up advanced computing ecosystems and and how to pull in a new technology and really try to give it the the best the best scenario for success and you know really how to bring it in. I think that would be from what I gathered from from you know the the time of the Munich Quantum ecosystem. And then I think I would say from Argonx, I've been here a year now, it's w the technology is advancing and so now I'm thinking a lot about ecosystem design, about how all of this technology, how these different modalities are going to come together. So not just the HPC part, but the HPC with multiple quantum then also, I mean, as you know, there's a lot of focus on on AI with the Genesis mission, what that's going to do to, you know, to revamping or to to redesigning maybe some of our compute ecosystem. And so now it's this great time. So I think it's it's constantly a a synthesis, I guess would be the overall arching theme throughout synthesis and kind of different flavors that I've seen throughout my career.
Daniel Rodan Legrain: I mean, this journey really, you know, put you at the center of one of the most interesting questions in science and computing today. I mean, how quantum systems really fit into this broader, you know, computing landscape. So, you know, for many years, as I was training myself, would even say quantum computing was really discussed almost as a standalone technology. And you've been one of the strongest advocates for viewing it as part of the larger computing ecosystem as you've very well said. So how do you see this relationship between HPC, AI and quantum evolving, let's say over the next decade?
Laura Schulz: Yeah, okay. So you know, again, I mean I I'm I'm always thinking about the end user, right? And and I maybe that's from my experience from being in an HPC center, but we've always provided resources so that, you know, very intelligent people who are cracking really hard problems and trying to gain insight, they've got the tools and the and and and the and the resources to be able to do that. So I'm always thinking about, you know, the end user and and the and the the success ability for science, right? The output for science results. So, you know, when I think about how all of this is gonna play together, I mean, again, we've got, you know, HPC that's done modeling and simulation, that, you know, where we've been able to to simulate the natural world and and and and get some results that we can actually quantify and study. When we look at AI, we're we're amassing a lot of data, finding pattern and process in that to try to give us some insights. You know, quantum I see as being able to simulate results or being able to simulate quantum mechanical effects on an on a system that is a quantum mechanical device so you can, you know, simulate it there. And so instead of doing you know, we we we we've done this for for decades too, right? We've simulated quantum mechanical elements, you know, in HPC, but again, that's an abstraction. Right. And so now where that's really important, we can offload that particular part of a problem, put it on to the quantum environment, get you know more accurate, ideally I'm running understanding much more accurate results for that particular part of the problem, bring it back into the overall simulation, the overall problem, and and pr and hopefully gain some new insights. So I mean, I constantly see these playing on the strengths of each other. in into kind of an assembling into an entire whole. That's how I see it.
Daniel Rodan Legrain: Perfect. So let me dig a little bit more there. mean, when some people hear about quantum HPC integration, mean, they often think about maybe plugging a quantum computer into a supercomputer. But in practice, and maybe a little bit more technically speaking, what does integration actually mean?
Laura Schulz: Yeah, yeah, that that's a really great question. I mean I always I always talk about these layers because an integration, again, this is not just unique to quantum to HPC. That this is this this story just keeps repeating itself throughout, right? the integration means several different things. One, you know, you actually have the physical hardware. So regardless of what's happening with the quantum and the amazing job that the physicists are doing with the with the qubit fidelities and and the control electronics and all of that, right? You you have to bring it into an environment where you have to couple it with the comp with the classical compute, right? And so, you know, the early systems that that we've been dealing with, you know, have been physically planted in the HPC centers and we expect that, you know, to continue in in various gradients as we move forward. But that literally means that you're taking what was a physics experiment, you know, lovingly cared for by you know PhD grad students and and technicians and people with physics degrees, bringing it into an ecosystem where we don't necessarily have those expertise and resources. But it's also the sensitivity of the system and the sensitivity of the system in an HPC environment where you've got, you know, very large systems running, different different facility infrastructures, all of that sort of thing. So you actually have to look about how you bring those systems in. And that and that includes everything from you know shielding it, and and and again, the the what the companies are doing to actually shield the systems, but then what the facilities' role is in in maintaining, making sure that those systems operate correctly. And that You know, that that's something that the companies can't completely do on their own. They do they do need the understanding of what the facility infrastructure and environment looks like. So you've got that point. you actually have like the connec the actual connectivity of the systems, right? So the the what sort of networking, the the actual architecture, where are you going to put like the common data? How are you gonna do the networking together? What is the what how are these gonna systems gonna communicate together? You've got the system software environment. That's kind of like the next layer up. And you know, this is non-trivial because now you've got quantum, which in HPC land we would call a disaggregated accelerator because it, you know, it has its own operational environment and it sits here. it's not fluid like what we can do with with CPUs and GPUs. it's a disaggregated accelerator, so you've got to deal with co-scheduling, and you typically tend to have you know, several different schedulers in in in accelerators like this. So you've got to coordinate that. You have to res manage the resource. You have to figure out the timing of when the job is on the HPC side and then parts of it go over to the quantum side. What happens to the to the jaw or to the the nodes that are on the HPC portion, what w you know the time that it takes for them to come back. You ha have to make sure that you're not keeping systems idle or, you know, or just waiting around for jobs. So how do you manage that? There's that whole level of of system software. That's even before you get to the application layer and before you get to the user culture. So there's all these different layers that you have to think about here and how all this works together. And then the other component, which is kind of a new component because we actually have systems launching now, is that now you've got systems distributed all across an ecosystem, whether that be regional or national or even international. What is going to be the coordination between those those quantum systems, between the quantum system and the HPC system that might be co-located or not? How does all that work together? So this is a really it's it's fascinating because there's several layers here and there's a whole bunch of questions that we're still actively researching at all of those layers. And Some of them depend on, you know, one layer, the answer is dependent on another layer's answer. And so you kinda have to to really think about how all these pieces come together. So integration is not just plugging in a system and, you know, plugging an Ethernet cable to it and and waiting for it to go. There there's a lot involved here.
Daniel Rodan Legrain: I can see a lot of open questions. So you mentioned all these layers. What would you say are the first layers that need to be integrated? you say facilities, scheduling, software, workflows, user access, something else?
Laura Schulz: Yeah, I mean that's that's a great question. And it's I hate I hate to be one of those, well, everything, yes, but but in some ways that that is true. I mean so like you know, early on, yes, we we had to get an understanding of the of you know, actually if a system was placed in in a system or in an ava HPC center, what sort of questions do we have? that's one component. But but already I I guess kind of answering this sorry very vaguely Every single layer has, you know, kind of critical steps that we have to take. And so to me it's kind of mixed. And and I think that that's actually important too because we're it's gonna be informing the other layers. And this is really when we talk about co-design, which is a word that, you know, that in in HPC space is is kind of our bedrock to to how we operate. When we look at the applications, when the applications team are thinking about how to put these systems together, how to how to utilize the compute from both of them, if it makes sense for their for their workflow for their application, they're gonna have particular sticking points. They're gonna identify bottlenecks, right? that's gonna influence what happens, like how you design the system software environment. you don't you you don't want you wanna make sure that you're developing for the user, you're not the developers aren't driving in and just telling the users here. figure out how to use it, right? And so there's this con as all of these lines and all of these people involved are are are working in these different spaces, they're going to end up hitting different parts of those layers. And so all of this is gonna work together. So it's a ri really long winded saying to way to say it all comes together at the same time, I think.
Daniel Rodan Legrain: And where are we today? mean, is it proof of concept integration? Are we in early production systems or are we somewhere in between?
Laura Schulz: So I we are we are in proof of concept, but a lot smarter than we were when we were doing proof of concept a year or two or three ago. so systems are actually, you know, coming into centers. And and this is not, I mean, this is for the the I mean I'm I'm really focusing now on this the quantum systems that are in the HPC centers. I mean, we you know, on the application layer and all, we've been doing a lot of discovery and insight. remotely accessing systems, you know, working on those, working with those vendors, doing all of that as the technology is developing. But I say would I would say we were proof of concept, but smarter than we were before. We're asking better questions. And you see that from the conferences. I mean, I was just at ISC a couple of weeks ago and I was honestly a bit stunned at at how much it was quantum this time. you know Four or five years ago when we held some of the first workshops in this space, to be really honest, we were kind of saying, Yeah, we're we're gonna need to, you know, figure out how to do this. And we were all honestly kind of slow blinking each other in in the sessions and and some of the papers that we had were it was very rudimentary. Every single conference, ISC, IEEE Quantum Week, SC, you know, you just constantly see more quantum and deeper. Deeper questions, deeper research, more people coming into the space. So I I would say that we are well on our way into the proof of concept, early, early deployment and all of the information and knowledge that that brings to us.
Daniel Rodan Legrain: Amazing. Well, I look forward to all these conferences coming our way. So Laura, from your perspective, what would you say have been the biggest surprises in trying to make that integration real throughout all these years?
Laura Schulz: I don't say I I wouldn't say it's a I I guess I wouldn't say it was a surprise, but it just reminded me of how hard we have to work. And and that is on the the translation and the and the communication. Right? So with with any project in general, I and project managers know this that basic I mean, there's a statistic from you know from your Pembok books that that like ninety percent of any project is is communication. You know, if you're you know, trying to to to coordinate and and develop things. And that's really true because we have such different communities that have you know kind of the starting point, with different cultures, different perspectives, different operation modes, di you know, different behaviors, bringing them together, but at the same time everybody is evolving and everybody's growing. So the, you know, the the quantum hardware companies They're actively working on their stuff, the the on their technology. The the software is trying to you know build this environment, the users are trying to figure out what to do. There's all these pieces, and because of the co-design part, the communication is paramount. And the translation of these different cultures and different needs and different user bases. And so translation, communication, coordination. that's always important. But because we have such a dynamic, fast-paced environment right now, that's even more important. so keeping everybody on the same page or at least with situational awareness of what's going on, and and and hopefully not letting factions silo too much into, you know, into their own solutions, right? That that then we have a hard time bringing together, you know, these technologies for the benefit of the user and their science. So I would I yeah, translation, communication, coordination, way more way more effort than than even I think a lot of us thought.
Daniel Rodan Legrain: I yeah, I guess that's a massive project that will really require a strong community coming together. So along these lines, Lara, I mean, you've mentioned all these layers that need to come together, but what do you see, let's say today, the biggest bottlenecks preventing this seamless quantum classical integration?
Laura Schulz: well for one, I mean, well, a few things. I mean, one, just the deployment of the systems, right? So we're we're s we're slowly having systems be deployed, but honestly, worldwide there's not that many that are connected, quantum systems connected with a through a s a a concrete through path with a system software environment to to an HPC system. So that we can do some of this work, right? And that and that's gonna come more as as time moves on. but that's part of it. You know, there's only certain places where this work, you know, can really be done, these hot spots. there's that part. And then of course, you know, in the the the big thing that the quantum technology itself is going to be going through a paradigm shift here very soon with you know with the emergence of fault tolerant computing. what that's gonna mean, you know, for the for the connection, the integration with with classical compute. There's the the application workflow stu flow there. Then there's also the error correction classical compute support of the quantum system and how those are going to how those are going to manifest within the same or different computing resources. but then but then you know How how are we going how are we going to manage the the various classical compute systems? So I think the rapid evolving of the the system at the quantum technology itself, but then also the the elephant in the room that's you know that's breaking the door down here soon with fault tolerance. And then what's gonna happen? With what we've already done. So NISC era was both good and bad, right? So there was a lot that we learned, but there's a lot that we're probably gonna have to unlearn, and that's just part and parcel. But then what can we reutilize? What can we learn from NISC and apply to fault tolerant in terms of again, the whole layers, the system software environment? Is can we take a whole lot? Maybe, hopefully. but in terms of application. So all of these layers coming up, we're gonna have a shift as we go to to fault tolerance.
Daniel Rodan Legrain: So maybe a little bit of a controversial one here, but do you think quantum systems will need to become more HPC-like or do you HPC centers need to become more quantum aware?
Laura Schulz: gosh. I I hate I hate saying both. I I really don't want to hedge my bets, but but but it is really true. you know, we when we when we think about, I mean, HPC supercomputers and that and that environment, these are very bespoke, well tuned systems, right? And they already have their own Operational demands, and I mean it's not like you just plug these things in and they work, right? I mean, there it's it's it's very mature technology, but it's continuously evolving itself, and so these are very balanced systems. So you can't just say, you know, quantum, you know, you guys just need to put on this sort of plug in for this, you know, this new plug in for this scheduler, or do this, or use a different schedule, or whatever. You can't have a unique solution. That works with your quantum system and then toss it over the fence to the HPC part, right? That that doesn't work. This all has to work really part in parcel. And so and so the quantum environment has to really understand, you know, what it takes to integrate to the HPC side, what those previous patterns and processes are that you know that it needs to abide by. It doesn't have to abide by all of them. You know, we we want we want some flexibility and innovation. But there are some guardrails that we need to respect. You know, on the other side, this quantum is just a completely new way of approaching computational results from a system. And we have to honor the physics of these systems and what they're going to require that's going to be delta to what we've what we've dealt with in the past. So there's gonna be learning and accommodation on both sides for all of this to work.
Daniel Rodan Legrain: Very fair, Laura. Let me switch gears a little bit. for most users, would say that the experience of a system is often defined less by the hardware itself and more by the tools, the abstractions, and the workflows around it. So let's talk about user experience. From your perspective, what are the biggest barriers to adopting these hybrid quantum HPC workflows today?
Laura Schulz: Yeah. Well, I mean, it's again the the just having the the actual ecosystem set up, you know, that that's something that we have to really work on. but then, you know, again, as I kinda mentioned before, you know, we're coming from from two different cultures. So one, you know, the HPC folks who have been doing this work and and and doing simulations of of they're gonna want to understand what what can I actually do with quantum. you know what what is the what is the actual behavior of the c n not even so much I don't want they don't want to know intimate details of the qubits right they want to understand you know behavior wise this you know when I decompose my problem this is what's happening at this point in the steps where can I push this off to quantum when does it make sense what's the benefit how is it actually beneficial over what I'm already doing Right. And so they there there's gonna have to be a lot of awareness building, decomposing the problems, understanding where the quantum fix fits in. That's that whole part there. So they're gonna have to get a lot smarter on the on understanding what the compute can actually do for them. on the on the quantum side of things, you know, to really understand, you know, where the classical can help with the pre and the post processing, where once once you've had the the power of the quantum compute and gotten what you can get out of you know out of the the run with that how can you bring it in bring it back to your overall problem how can that system support you for for your quantum for your quantum calculations so it's it's it's funny at at ISC we had a a a boff on on workforce development and what was interesting to me I asked the question to this group and I said you know the physicist raise your hands what do you want to know most, you know, moving forward about like integration? And all of them said in some flavor or another, more understanding of what it is HPC is and and you know, what it means and and and how I'm supposed to think about it and what I can do. And then reversely, the HPC post were like, Yeah, we need to understand more of the quantum side of things. And right now we're at the point that they actually have to get get get cozy with the qubits and they need to understand kind of hardware at a level that we wouldn't ask them to understand if they were in the HPC space. You know, right right now, we have so many abstractions built up in the HPC realm that we try to we remove the users from intimate understanding of of the bits and the transistors and all of that, right? But on the other hand, right now, because of where we are with quantum, people who use quantum kind of need to understand this and have a you know have an understanding of how to operate it. So what we want to try to do is remove that away from them as much as possible and build an environment that they understand they understand enough of how the compute works and behaves, but we don't want them to have to be quantum experts in this space in order to be able to use it. That's these folks are experts, but that that's that's quite a big ask and that's that's quite a big bridge to cross.
Daniel Rodan Legrain: Indeed, yes. And how do we give users, let's say, a common experience across different quantum technologies altogether? I superconducting qubits, neutral atoms, trot ions, spin qubits, and others.
Laura Schulz: Again, a fantastic question. I mean, I think the first thing is literally just the access and the you know the democratization of the access. So hands-on is the most powerful thing that we can do. and and already we're already starting to talk about infrastructure. how how do we how do we provide access for these systems on a common back plane that also has, you know, at the front end it has the same software environment, right? And that they can utilize. regardless of the technology that's common enough, you know, to get them to a certain level, but then it also honors the unique physics of the individual systems when that part becomes important, if if that makes sense. So it's you know, it it's about making sure that we we get the access, but that we're providing the ecosystem that they can use it. I mean, what what's happening, and this is this is completely normal for for early days of technology. We saw this in HPC. We we, you know, this is a reoccurring pattern where you you know have technologies that come up and and everybody wants to build a full stack solution, right? From the hardware to how it operates to the application software that it uses. And but what you end up getting is a whole bunch of different users on one side, a whole bunch of different devices on the other, and then you get this Hornets nest of workflows and tools and all that aren't really all working together. And the problem that that causes is that you're gonna end up getting a silo effect or you're gonna end up getting you end up getting necessary allegiance to one technology and one, you know, one s series of resources and a workflow because that's where there's been the investment, either in your center or you know, from your chair or whatever it is, right? So you end up kind of get locked in to particular areas and maybe that's not the best for your Applications for your work. We don't want to do that. So kind of the role that we have as HPC centers and you know and also would say as governments is to figure out a way to have to provide access to these different resources, but to provide kind of as much of a common middle ground with that software environment, you know, coupled with the application side so that these users can experiment. And be able to understand the best technology for their particular problems.
Daniel Rodan Legrain: Fantastic. Let me switch gears a little bit, again, and let's talk about ecosystems and why they matter. mean, you've worked deeply in both the US and the European advanced computing ecosystems. So how would you compare their approaches to quantum HPC integration?
Laura Schulz: that's really a great question. lots of great questions, Danny. on so on the on the European side, you know, it it started off with and because of the you know the really impressive efforts of like Euro HPC started off early with providing quantum hardware startup companies, you know, with with first first system acquisitions. Putting them into HPC centers, really with the mandate early on. Here's a quantum system. Figure out the technology is still rapidly evolving, but go ahead and start figuring out how this is gonna be c part c become part of the computing ecosystem. Right? And and and so yes, it's still evolving, but go ahead and start connecting it. figure out figure out what you could do. So this really this really facilities level, ecosystem level, hands-on, you know, direction. the the US side is and the reason that I you know love DOE so much is is because of this big mission directed science, right? And so really looking at of the applications and things that we that we have a priority To work on where does quantum come in? So I think it's coming heavily from I think it's coming from the the the HPC center side. I would say the US side was coming heavily from the application side. What can we do with this? And really trying to experiment under the hood and testing, you know, virtual systems first or cloud-based systems. what we could do on the application side, but then And then on the European side it was very hands-on with the technology, get it into a center and start building from there. So I don't want to say top down and bottom up, that's that's not correct. But it was kinda but it I do kind of feel like there was two ways of looking at it, and I think they're fully complementary and and I and I do think that it works very nicely together.
Daniel Rodan Legrain: Fascinating. So how can the US and Europe, let's say, learn from each other as they build quantum H-15 infrastructure?
Laura Schulz: Yeah, I mean I again that they're gonna come from they're gonna come from different perspectives, but at the same time we're also, as we kinda mentioned before, I think the the beauty of this of the of quantum HP C of this of this little hybrid environment that we're building up is that we all are working together really well right now. I I really enjoy this space because I think it is very collaborative and and cohesive. so I think that the you know what What the US side and these big, you know, the big companies that are really pushing forward and have some great results, the the the focus on applications, you know, what they're gonna be bringing to this space and what the the European and also we can't forget the the Japanese space, which is is is really amazing and and Asia in general is really is really kicking up in this space. Everybody's coming at this from a slightly different starting point, a different vantage point, and and maybe a different strategy. And so Bringing those together, yeah, th this this is really powerful and and and very useful. and we're going to learn from each other and you know we're gonna do things differently, and that's okay. What we wanna try to do is we want to try to we don't you know we talk about standards too much. We don't want to talk standards right now, it's all still too new, but we wanna talk interfaces. we wanna talk the ability of where are there particular couple, you know, coupling points, where can you connect in To this overall environment that we're building up with the software, with the hardware design or ecosystem design, connectivity, all this. Where can we find certain points that it's easier if you just kind of follow this particular interface or contract, be a simple API or a more advanced, you know, understanding of that? Where can you kind of connect into this so that we can have multiple different approaches, but still not silo? Right. I think that's gonna be really important. And and gl I think quantum is just very global. I mean, I mean, y you know this. I mean, QBlox you know, is in Europe, but it's in the US. It's working across like the entire global ecosystem. Pretty much all companies do this in some form or another. So I think everybody's a part of these multiple ecosystems. And so I and that that's also very encouraging that this is all going to eventually work together very nicely.
Daniel Rodan Legrain: amazing. And let's talk about the people because none of this works without the people. mean, throughout your career, you've really built all kind of multidisciplinary teams spanning from what I heard, physicists, computer scientists, software engineers, mathematicians, system architects, you know, all kind of user facing teams. And I'm sure many other people coming together. So what have you learned about, you know, leading teams where, you know, people often speak very different technical languages? especially in such a complex space.
Laura Schulz: Yeah. Yeah. Yeah. So I mean the thing is is like we don't want to we don't want to generalize everybody out, right? You so you need you still need those experts in each and every space. the the team that I was privileged to lead before was was exactly that. It was this group of experimental and theoretical physicists, computer scientists, software engineers, you know, y developers, electrical engineers, you name it. And everybody's expertise, you know, came into that in c into that pool. You know, I mean we it's not a melting pot, I think it's a salad, right? And so, you know, these people have all their different unique backgrounds. But then again, it it comes down to spending a lot of time talking to each other, understanding each other's perspectives. And you know, and we see this I mean this wasn't just like my the like an individual team at a at an HPC center. There's an initiative led by Amir Shahada in at Oak Ridge called OpenQSE, right? And that and these, you know, working like bringing all of us together in that ecosystem. And again, the same sort of thing, you know, from the hardware companies, you know, software developers, et cetera, control electronics companies, really trying to sit down and and build common interfaces, common, you know, points where we can all agree that you know our technology should do this at this point and and that we should be able to work together. So it's I and I also think there's gonna be a really interesting really interesting for folks in the future when we talk about workforce development because we still need those experts. You know I mean you still need people that we can go to to say this person knows about you know about fidelities with the qubits or or knows what happens, you know, for the with its error correction. At the same time I think everybody is going to have to have the ability to zoom out of their prospective expertise and have an understanding of what's going on, you know, in other spaces. And and not this is not just applied to HPC and quantum. I mean, I can tell you, sitting in on on discussions with with AI, you know, and and building up the AI ecosystem, there's a lot of work. that's already being done between HPC and AI integration and what that means. And and there's already big investments happening in there. And so there's already processes being built up to solve that particular, you know, connection problem or or whatever those issues are. And we need to be able as a quantum community to understand what's going on there, leverage what we can again for this whole ecosystem You know, again, for the benefit of science, we want to have end-to-end workflows. So we need to be paying attention what's happening there. And then at the same time, we need, you know, the AI and the HPC people to be actively in this conversation. So I guess one of the biggest things and that I always say is when you go to a conference that's like SC or ISC, you know, go to the quantum things for sure, but go spend some time over in some of those, you know, pure AI and pure HPC. Because I I guarantee you you're you're going to hear things that resonate with what you're with what our community is working on for sure.
Daniel Rodan Legrain: Duly noted, Laura. I see quite a lot of questions popping up in the chat. So let me maybe end this conversation from my end with a broader future looking question, and then we can switch to the audience questions. You describe the future of computing as increasingly heterogeneous. So if we sit here again 10 years from now, what does state of the art computing look like by then?
Laura Schulz: if I had my wish, first of all, I mean, fault tolerant is is you know gonna be, you know, moving forward, we're going to have those first systems. I mean, for me, we would have more systems with HPC connected to it. we would have a a deeper understanding, we would have a more evolved and matured software environment. But kind of all of it comes down to there would be some really interesting use cases. There would be some really interesting applications that people have sorted out that they did by asking questions they couldn't have asked before because you know the environment wasn't set up for them completely. So you know we we had some early, some really just interesting, early, very early, understandably, you know, results from when we hooked up the first quantum computer to the Supercomputing Center in Munich. Where I worked, right? The system. And rudimentary, rudimentary software, completely understandable. However, it already changed the way that the questions were being asked and ideas were being formulated and thoughts about the workflow. So the more technology we set up, the more infrastructure or ecosystem that we set up, the more mature it comes, the more interesting questions and results that are going to happen. So I just want to see really great science.
Daniel Rodan Legrain: Fantastic. Let's look over to a great science. Thank you very much, Laura. That was a really fascinating conversation. Let me bring in a few questions from the audience in the last few minutes. I see quite a lot of questions. Let me start by timeline one. So will hybrid quantum classical cloud services become more common in 2026, meaning now to 2030? or is it still a little bit too optimistic?
Laura Schulz: Really good question. so I mean w when when we say cloud, we just mean a facility that would have access to a quantum system and access to an HPC system and have the capability to put them together. I mean, I think in terms of very rudimentary integration, that's I think that that's achievable, right? when we talk about more nuanced and more mature and complex integration. That's still gonna take a little bit of a while for us to do. One, because we're still learning how to do it. But but yeah, I don't think it would be I don't think it would be out of order by any means to start seeing you know, some sort of access to rudimentary HPC and cloud compute on on on cloud services.
Daniel Rodan Legrain: Thank you very much. Here's another one. You've mentioned it's important for quantum and HPC communities to really understand each other's technologies. What are some good resources for quantum physicists to better understand about HPC? And if I may add a second part of the question here, what do you think new generations of scientists that want to enter the space should be learning and focusing on?
Laura Schulz: This is just gonna be a contact sport, to be really honest. So I think you know, there's been a lot of educational resources built up for kind of quantum-only understanding, you know, so how to do this. What what I do think that we need to see more of coming up is quantum, quantum explained for HPC practitioners, right? And and kind of taking N let's not start completely with the concept of you know of of qubit and superposition and all these. Let's let's really look at it, you know, when you're when you're dealing with you know with your code and you're dealing from your C or your C environment and and and how you're thinking about you know your your your code decomposition and and everything that's happening. This is this is literally the the computational behavior that quantum offers. This is how it operates, this is what you can do. So there needs to be how I say it this right way. There needs to be HPC coded language on quantum, right? I think that's kind kind of the big part. that needs to happen. And and and I'm hoping that that we start having more educational courses from that direction, taught from the HPC perspective, what quantum means, not so much from the quantum community trying to explain, you know, to the HPC. Does that make sense? it's coming from that aside. I think that's gonna be really important. So I'm I'm hoping that that a lot of folks kind of pick up that mantle and and and drive on that pretty hard. on on the quantum side, I mean the the the the kicker is you know in the physics community there are a lot of people that are that are that ha that are using HPC, right? So I mean, you know, you've got the quantum chemists and and you've got these folks who they've been utilizing simulation on HPC systems to understand, you know, some of their quantum mechanical effects. So they already so there's a lot of folks that already have that knowledge, right? And so now they're basically being able to take it, take part of that problem onto a quantum system. There's that class. For the pure, pure physicist that that haven't done the HPC or haven't really done the that sort of computational simulation part, yeah, that's a great question. I I wanna say very blithely that it almost seems like you have to just start with HPC, but yeah, I might have to think about that. Like how how do we onboard the the the physics folks for the HPC that maybe don't have the understanding of HP C simulation? I'll I'll have to get back to you on that one.
Daniel Rodan Legrain: Sounds great. There's actually a similar question, but in the other direction. How do we get people from the traditional HPC industry to really understand which algorithms are, let's say, most effectively suited for quantum?
Laura Schulz: Yeah, I mean this this again, and and I mean maybe I'm coming at this kind of naively, but I mean g once you start putting quantum on anything, you know, there's there's this like, my goodness, you know, this is gonna be so hard. And part of that is because again, it's been wrapped up in the in the culture you know, of very talented physicists who understand, you know, who understand the lingo, who understand all these components, and there's a translation problem to the to the HPC folks. So to me, it always comes down to behaviors. What I would like to understand is like, for example, in in HPC space, I mean we we do have HPC algorithms, right? And we have HPC algorithms that can that have perform a particular function that can be, I don't want to say reply replaced, but that we can we can take HPC versions of the work that the algorithms were doing there and and utilize quantum algorithms in their place, right? So I guess one thing would be, let me put it this way, I think probably the best thing to do is to get the HPC algorithm developers and the quantum algorithm to get developers, put them in in a really nice location for a long period of time and lock the doors. because I th I think I think that there's gonna be a need to be a lot of sitting down and figuring out at that level to be able to understand, you know, to to understand what the potential would be.
Daniel Rodan Legrain: I like the idea.
Laura Schulz: But but but it'll be really good it'll be really good pizza, you know, the food and it'll be really nice. But I again this just comes with the idea that it just has to we have to we have to we really have to put these communities together. E we're doing a great job, folks. Everyone's doing a really nice job of all of this. We have to do it even more so as we move forward.
Daniel Rodan Legrain: Fantastic. Let's do it. Let's talk to the people that have the power to make it happen. And I think you're sitting among them and you're part of them. Laura, I think we have time for one more question. So do HPC professionals and researchers like to collaborate with full stack quantum computing companies or any other companies in the value chain like software, QP, developers, control electronics? From the HPC point of view, who is the easiest or more practical to work with?
Laura Schulz: so Well, I mean I I think I I I'll give a shout out to integrators in this space now. So you know, when like again, HPC systems are are very complex entities, right? I mean these these supercomputers are an acc an an accumulation of of you know various network technologies, of of data store technologies. of the not just the you know the the CPUs the GPUs the telemetry environments the the the power part all these are all pieces that come together to build a supercomputer right and and companies who do this sort of thing are are integrators who can pull all these different vendors together all these different solutions work with you know the HPC centers to figure out what the best solution is and then and then work with them to map out what that solution is going to look like and then to and then to actually operate it. So I do see a big role for integrators with quantum companies because they're going to be the ones that are also going to look at the various technologies, the various solutions, again trying to you know work with the the the clients at the end what is the solution that's needed and then how are those components gonna fit in. So I mean I think integrators are a are a big one. And then there's the there's like the hardware integrators and the solution integrators, but then there's also, you know, software integrators. you know, so I mean again like one of the examples, not not a shock, would be, you know, something like NVIDIA and some of the stuff that they're doing with quant because they understand the software environment from the quantum and the HPC. Right. So it's again anybody who is dealing with Thinking about all of this in whole and not as individual, that's gonna be the sort of companies. And again, there's also quantum companies that are thinking that way as well, right? That are moving over. So I guess it it's interesting, it's not really a one or the other, it's more a philosophy and approach to how the company is thinking about this. Just buying a generic or just buying a quantum system that just works as a quantum system. that puts a lot of burden on the on the maintainers and the operators who are purchasing it, right? So those who are actually working on the hybrid solution are are the ones that we want to that we want to encourage.
Daniel Rodan Legrain: Fantastic. We definitely need more integrators, Laura. Laura, thank you so, so much for joining us today. mean, your perspective is really valuable. You've helped really shape this field from all kinds of different angles, national labs, superconducting centers, European initiatives, quantum HPC integration as a whole, user facilities, industry collaborations, and all kinds of international programs. One of the themes that really stood out throughout this conversation is that the future of computing will not be built through one single technology alone, and there's many layers to think through. It will be definitely be built by bringing together all kinds of different technologies, but also different disciplines and different communities, quantum, HPC, AI, hardware, software, facilities, users, you name it, many others.
Laura Schulz: Yep. Yep.
Daniel Rodan Legrain: That, I would say, connects very naturally to what we often say at QBlocks, building quantum together. It's not just a motto. It's a recognition that useful quantum systems will require a whole ecosystem. And I think today's conversation has been a fantastic reminder of that and how it looked like in practice. As a small token of appreciation, Laura, we've sent you a QBlocks mug with our motto, building quantum together. Hopefully, it accompanies you on the next round of strategic planning at Argonne. To everyone that joined us today, thank you very much for spending the hour with us. If you join later or want to revisit the discussion, the recording will be available on the QBlocks YouTube channel and the audio version will also be available on Spotify and Apple Podcasts as part of the Quantum Builder series. Stay tuned for more exciting Fireside chats coming soon. And Laura, thank you again and thank you everyone for joining us today.
Laura Schulz: Thank you so much for having me, Donnie, and the Q Blocks team. Thank you.
Daniel Rodan Legrain: Our pleasure, Laura.
