GSA Awards Celebration 2025 | Pat Gelsinger, Snowcap Compute

Clips
More from GSA Awards Celebration 2025

Pat Gelsinger

Chairman of the Board

Snowcap Compute

play_circle_outline Pat Gelsinger's insights on current trends in semiconductors and AI.

play_circle_outline Transforming Semiconductor Innovation: The Role of xLight Technology, Advanced Materials, and 3D Packaging in Moore's Law

Info
Transcript

Pat Gelsinger, Snowcap Compute

Pat Gelsinger

Chairman of the Board Snowcap Compute

In this interview from the GSA Awards Celebration, chairman of the board of xLight Pat Gelsinger joins theCUBE’s John Furrier to discuss the resurgence of the semiconductor industry and the critical role of new materials in driving future innovation. Gelsinger details the recent announcement of xLight, a next-generation light source designed to revolutionize lithography by moving beyond current EUV limitations to wavelengths of two nanometers or below. He explains how this breakthrough addresses the economics of Moore's Law, allowing for faster wafer producti... Read more

explore Keep Exploring

What developments and innovations are contributing to the current momentum in the semiconductor industry? add

What advancements or changes are necessary in the semiconductor industry to sustain its growth and improve economic viability in light of challenges to Moore's Law? add

bolt Powered by CUBE AI

Pat Gelsinger, Snowcap Compute

search

>> Welcome back onto theCUBE here at the San Jose Convention for the Gala Award show. This is the GSA celebration. It's the annual award for it. It's the Oscars of the semiconductor industry. Pat Gelsinger, CUBE alumni is here. Always has his finger on the pulse. Your DNA, Pat, goes way back. Welcome back to theCUBE. This is not red carpet. There's no red carpet, but it's kind of a red carpet.

>> Indeed. Indeed.

>> Thanks to coming on.

>> Well, hey, I heard you were going to be here tonight and I just had to come.

>> Word got out that theCUBE's here and you had to just take your private jet in and come right in. Tell me what's going on because we've seen so much action in semis. And again, I think I've interviewed back 20 years ago when I first did a podcast with you at Intel, and then now so much has changed. What are you working on right now? I saw your interview on Yahoo Finance. You got some cool things happening.

>> Yeah. Yeah. Well, I think semiconductors are hot again. And clearly AI, the whole surge, but it's really all of the new materials that are occurring. We have cryogenics and quantum computing and so much going on, which fundamentally it all runs in semiconductors and some sort. And this week we announced xLight and xLight to next generation light source. And if you think about a fab, what's in the middle of it, lithography. What's in the middle of lithography? Light. And God said, "Let there be light."

>> Yeah. And the physics, the speed of light, we'll see that. The photonics is a hot area. We're seeing all these new interconnects. Take us through in your vision, what needs to happen for semis to kind of continue the thundering pace it needs to go to give the connectivity, to get the horsepower?

>> Well, clearly what we've done this week with xLight to me is an important piece because when you think about Moore's law, it was always three things, John. It was always more transistors, better transistors and cheaper transistors. But as we went through the EUV transition, the transistors are no longer cheaper. We lost the economics of Moore's law. So what we're doing with xLight is being able to not just be 13.5 nanometers where EUV is today, but be able to go to two nanometers or below in wavelength of light, really spectacular, but also has improved the economics. Higher power light allows us to make the wafers faster, which allows us to improve the economics, get the economics back into Moore's law. So this has been an important week, but also we see that, hey, 3D packaging is more important. New material types get to be more important. We need to move from copper to optical for all interconnect as well. So it really is a time, as we would say, we're rebuilding the entire technology stack. There's never been a better time in tech.

>> And the demand is massive. Okay. It feels like the PC revolution where, okay, the 286, 386, that generation is now happening at larger, larger scale. And if you factor in geopolitical pressure, the incentives... Take us through what the new normal is because most of the lay people, even now in the mainstream, AI infrastructure is gone mainstream. I mean, people are talking about chips and GPUs and XBUs.

>> Ain't it cool?

>> I've been doing that... I talk about this for like 20, 30 years. So explain the light piece and why that's important and what that will be. What does that provide the value to? Explain to the end user who's like figuring out, "I got an iPhone, I'm going to have a body device. I'm going to have a hyper-converged edge soon with wireless spectrum, unlicensed coming together." You're going to have smarter devices at the edge. What's all this mean for... How does that get better?

>> Yeah. Well, fundamentally, if we go back to the basics of Moore's Law, put more transistors on a chip, make those chips smaller, lower the power consumption of the chips. And that's why, hey, in your pocket, you now now carry what used to take a whole data center to do. It really is pretty spectacular. But can we keep making things smaller? Can we keep lowering their power? Can we keep improving their economics? And to do that, lithography has played a central role for the 50 years of the semiconductor industry. So how do we make it better? We have to attack the hardest problem, light. And making smaller wave light allows us to make the wires smaller, the transistors smaller, and improve the cost of it.

>> And if you were back in school as undergraduate, maybe imagine our PhD program, knowing what you know now, what you're working on, what science is important now to study?

>> Well, I think all of the hard sciences continue... They're back in vogue, right? We sort of had a decade where, hey, if you weren't a software CS guy, right? Well, now all of a sudden everything is going to enable us to solve problems in material science, biology, chemistry, physics, all of those become back to the center of what we can do for the future. So it really is to me, it's back to the basics, baby.

>> Yeah. Back to the basics. Physics, material science, engineering.

>> And I call it the trinity of computing, right? Where we're going to have classical computing, like we've been building and CPUs. We're going to have AI, which is that scale processing of AI models, large language models, but we're also then going to have the third leg, quantum computing. Where there's so many algorithms that are not computable in digital space that are only going to be computable in quantum space. And as that materializes over the next couple of years, okay, almost everything that we can write in the equation for will now become computable.

>> The Holy Trinity of computing laid out by Pat Gelsinger.

>> Amen. You got it.

>> All right. So I have to ask about quantum computing because I love quantum. There are use cases where quantum is doing very, very well right now. We see some visibility on that, but also there's conversations around that we're having around how GPUs, XPUs and CPUs are contributing to the advancements of quantum. Can you share your vision on how the computing world today can lead into advancements of quantum?

>> Well, when you think about many of these algorithms, like if we think about the DFT algorithms or so on, there's a lot of pieces of that that are very data intensive that are computable in classical terms or digital terms and big data sets and all of that is great, but there's core pieces of the algorithm that you have extraordinary NP complete or exponential computational requirements that literally on a classical computer would take billions of years. And those are the things that now can be solved with quantum, but you combine it with these extremely large data sets that were now computing at scale with AI. These large HPC-

>> AI is like the test kitchen for quantum.

>> Yeah. It's a warmup. It's a warmup.

>> .

>> So to me, when all three of those come together, it'll be the greatest time in technology, the greatest time in science history. My only regret is I'm not 20 years younger.

>> I know. I always say that all the time on theCUBE. I wish we could be 20 years old. We'd be really having fun. Pat, I have to ask you, since we're at the GSA Awards, this is like the Oscars of semiconductors. You had an amazing career in it and continue to do that. What's changed so much? This is like a Hollywood event. theCUBE's here. It's quite the glamorous scene. What's it been like over the years, this event? There's a bunch of semiconductor nerds hanging out, getting awards for each other. Now it's like that-

>> Well, I do think there's sort of like the Nobel Prize, right? Where people have always said, "Oh, it's really important that we have that highest of achievement being recognized for." Well, of course we used to just say, "Hey, we're going to have the Oscars, right? We're going to have the Golden Globes and all that kind of stuff." And that's a bunch of actors patting themselves in the back, right? And so they haven't created value. What are they? They're actors, right? They're pretending. These are people that are creating real value for humanity, real technologies, real manufacturing that changes what humans do in their life. And to me, this is so much more important. If you invited me to the Oscars and gave me a front row seat, I'd say, "Do I really have to go"

>> I'll watch it on TV.

>> When you come here, you're honoring people that have done things that truly improve the lives of humans.

>> Their life's work, invention, building. And if you look at who's in this room, the world economy is here.

>> Yeah, indeed. And increasingly, as you think about a digital AI driven world, what we do here truly is the economics of the planet because everything goes digital, everything uses AI, everything will come computable with quantum. Truly, the economy of the world rests here.

>> It encapsulates Silicon Valley in a nutshell. Pat, always great to see you.

>> Thank you so much.

>> You're awesome.

>> Sorry to intrude on your program today. And Vellante, where are you, man?

>> Dave, Dave Vellante. Pat Gelsinger is calling you out here on theCUBE. Let's go. What's happening? Pat Gelsinger, a good friend in theCUBE. Always a pleasure. Thank you so much.

>> Thank you.

>> The legends are here. And also, Pat is not out of the game. He is back in the arena. He's never left. Making it happen. This is a place where people who are changing the world or keep changing it, are here celebrating. But once this celebration is over, the game continues tomorrow. And 2026 will be another fantastic year for AI infrastructure. I'm John Furrier, host of theCUBE. Thanks for watching.