 So we hear the supercomputing conference and who are you? So I am and I work out of Sandia National Labs. So what is the Sandia National Labs? Well, we're a DOE facility based in New Mexico, the United States, high performance computing and lots of engineering problems. So do you work on all kinds of what the future of supercomputing is going to be about? Yeah, so we're busy working on DOE's exascale vision at the moment, so that's technology that should be around in the early 2020s. Exascale? Exascale, yeah. That's like a vision people would like to reach some kind of, what does it mean? Because it's a thousand times faster than pedascale. So that's loads and loads and loads of calculations every second and most of that means just really fast simulations, so a thousand times, think about a thousand times faster to solve the same science problems. But aren't some people saying this is not going to happen? Some people are saying that, but we're working our hardest to make it happen and we're pretty confident we're going to get there. So how fast is the fastest supercomputer right now? It's about 20 petaflops, so we've got about 50x to go. So we've got, it's a long way to go. So 20, like is it the Chinese one? Is it 20? Chinese, well the largest DOE is 20. What is DOE? The Department of Energy, so that's the largest one. American one? For us it's 20, it's American one. Yeah, in China there are 20. In China there are about 100, so they've only got 10 times to go. And so you're going to do 50 times faster in the next five years? Hopefully before then. Before, so that's faster than Moore's law, right? What do you call it? Sandia's law, what do you call it? Yeah, faster than Moore's law, yeah. So is it because you don't focus on floating point as much, you don't care about the floating point, you're doing all kinds of other specialized processing or there's nothing to do with that? How do you architect the supercomputer? Sure, well there's a lot that goes into that, right? So we think a lot about our applications and our workload and then we're trying to make all of the components that that needs, so that's memory, that's processor, that's networking. We need to make all of those go faster because if we can only make one of them go faster, we'll get a bit of an improvement but we won't get to where we need to be. So it's a question about taking all of those pieces, making them all go faster together so it's balanced, it's not balanced, one word. So you synchronize the technology somehow. It's not good if one goes too fast. Yeah, we're looking for improvements in every area. So Fender's are really good at doing one piece of that. We need to do memory and we've got to do networking as well. So we've got to work right across the board with everybody. So you're building something that has the biggest memory bandwidth in the world? That's what I'm hoping for. So how big is it going to be? I can't give you real numbers yet but it's going to be big, so think about it. We're looking for 50 times faster than the machines we've got today and we'd like to do better than that. And it's going to fit in a small data center? Where's it going to fit? No, these are big data centers. It's going to be big like this? Size of a football pitch. Football pitch, oh yeah, there's a Broncos game right now. Yep, well we could take them all off and put our supercomputer down instead. All right, so here at a supercomputing conference, do you think it's going to be exciting what people are showing, talking about here? What's going on at this conference? We're expecting some new machine announcements this week. There should be some new announcements on hardware. Can't talk about what those are yet, but basically start tomorrow on the show floor. Pretty exciting stuff. There'll be some new machines in the top 500, new machines on HPCG, which is a benchmark. So we're pretty excited, it's going to be a busy week. And how many people work together with you in that center where you are? So we've got about a hundred and 200 people working on that stuff at Sandia, but there's other DOE labs, six DOE labs involved in scale across the country, so think about it, like 1,000 people. 1,000 people working on the future of HPC, the future of supercomputing. Yep, you got it. And that's important for everything in society, right? Health, everything. And if you think about these big machines, right? A big machine from 10 years ago is now at the universities and the students are using that to solve problems. So everything we do at the high end eventually ends up, you know, businesses and industry and designing the next aircraft and all these kind of things. So what's going on? It ends up in the smartphone. Yeah, it does. So the smartphones right now, if you go back like 25 years, a smartphone today is faster than that supercomputer.