New Oregon State supercomputer will be like ‘a time machine,’ Nvidia CEO Jesen Huang says (Q&A)

Oregon State University announced Friday that is plans to build a $200 million supercomputing research center at its Corvallis campus, an innovation hub to advance climate science, materials science and oceanography.

The supercomputer that anchors the facility will use artificial intelligence to conduct advanced simulations of the real world. Nvidia founder and CEO Jensen Huang and his wife, Lori, are helping fund the project with a $50 million donation. They met at Oregon State as engineering undergraduates.

In connection with the announcement, Huang spoke to reporters last week about his hopes for the project and the future of computing. Here is that conversation, edited for length and clarity.

Jensen Huang: The background of it is that Lori and I both went to Oregon State. We’re proud alumnis. We met there. We have a great fondness for the school and, we’re also proud of the school.

As you know, Oregon State is a world leader in many, many fields of science that are very important to our future: environmental sciences, ocean sciences, material sciences and robotics.

And for the very first time the computer industry has the necessary technology that might be able to help these researchers advance these important fields.

And that technology is called artificial intelligence. And so we’re going to help Oregon State build quite a large instrument, otherwise known as an AI (artificial intelligence) supercomputer, that helps their researchers to advance their science. And so, so that’s really the background of it.

My sense is the idea is to take some things that Oregon State does especially well, and take something Nvidia brings to bear with artificial intelligence, and advance these particular fields.

Yeah, that’s right. And of course, because Oregon State is so passionate about those fields of science and so good at it, it attracts a lot of researchers and students who love the same thing. And these particular fields of science are really important for the future.

We need institutions like Oregon State to get our arms around the best ways to mitigate climate change and the best ways to come up with strategies or new technologies that might slow it.

A lot of researchers that are there need a new type of instrument, a new type of technology that help them do their work. And artificial intelligence is that thing.

We’ve been talking to Oregon State about something like this for probably five years.…How best can we make a contribution? How can we inspire researchers there to take advantage of the capabilities of artificial intelligence? How do we develop the next generation (of) students that can utilize this capability?

And you also need a place to put it. You need a collaboration center where scientists of different disciplines can come together.

Climate science is a multidisciplinary science. It’s one of the most complicated fields of science. And so you need to create the environment, the conditions by which multi domains of science and students…can come together in one place.

It took until now to really pull it together.

What message would you want to send to other leaders in technology about AI and the importance of higher education’s access to this technology?

Artificial intelligence is one of the most transformative technologies that the world’s ever known. We can apply intelligence to problems at an extraordinary scale. Humans have great intelligence, but we can only … wrap that intelligence around so much data.

Artificial intelligence, especially with today’s computing scale, could solve problems that no humans could possibly imagine wrapping their arms around or their head around.

This instrument’s available for the world’s largest technology companies that apply it to all kinds of very important problems like shopping and music recommendation and things like that.

But we need to put this technology in the hands of scientists so they can apply it to the most important challenges. Most universities today still (haven’t) come to grips with the idea that in order to advance the most important fields of science, you need a new type of instrument.

Could you tell me a little bit about how (the donation) might help support the chip industry in Oregon and beyond.

Oregon State is really dedicated to advancing material sciences. And the semiconductor industry is near the limit. It’s near the limit in the sense that we can keep shrinking transistors but we can’t shrink atoms.

Until we discover the same Pym particle that Ant Man discovered (in Marvel Comics) our transistors are going to find (their) limit. And we’re at atomic scales. And so this is a place where materials science is really going come in handy.

A great deal of the (semiconductor) industry is going to be governed by the advances in materials sciences. And material sciences today is such an enormously complicated problem because things are so small.

Without a technology like artificial intelligence, we’re simply not going to be able to simulate the complicated combination of physics and chemistry that is happening inside these devices.

Artificial intelligence has been proven to be very effective in advancing battery design. It has contributed to advancing more durable and lighter weight materials. And there’s no question in my mind it’s going to make a contribution in advancing semiconductive physics.

There was some back and forth with you and your counterpart at Intel (Intel CEO Pat Gelsinger) about the status of Moore’s law and whether it is in fact dead. Let me ask you about that. Materials science, can it extend Moore’s law? Is Moore’s law permanently dead, or could there be advanced that would revive it? (Moore’s law is a prediction by Intel co-founder Gordon Moore that computing power would increase exponentially even as the cost of computing fell.)

When something dies, it might be reincarnated, but it dies.

The question is, what’s the definition of Moore’s law? Moore’s law was about doubling performance every year and a half, or reducing cost by two (times) every year and a half.

Dennard scaling (a maxim that states as transistors get smaller their power use falls, too) ended close to 10 years ago. And you could see the curves flatten.

And so the ability for us to scale 10 times every five years is behind us.

We could keep our head in the sand, but we have to acknowledge the fact that we have to do something different. That’s really what it’s about. If we don’t do something different, and if we don’t apply a different way of computing, then what’s going happen is the world’s data centers are going to continue to consume more and more of the world’s total power.

And so this is an imperative. It’s an imperative that we change the way we compute. There’s no question about it.

It’s harder to find chip designers or people just interested in the material side of semiconductors or the manufacturing side. I’m curious to what extent that factored into your decision to make this donation. This is a pretty exciting thing for younger people or college students to be able to have access to.

Well, the researchers at Oregon State are going to experience computing like it’s a time machine. And the reason for that is because something that used to take a month now takes a day.

That’s a time machine. And you can see the future like you can’t possibly imagine.

So researchers that are doing climate science, researchers who are doing oceanic science, materials science, are about to get into a time machine. The computer science behind it is so magical.

So now we’re going have this instrument stood up at Oregon State, it’s going to inspire a whole bunch of new students that come and use it to come and build the next generation of work.

As Oregon positions itself to try to cash in on the CHIPS Act and revitalize our semiconductor industry, one of the things that we have been very conscious of is the lack of a major engineering school. Do you have those aspirations for Oregon State?

Climate science, materials science, robotics, are going to be some of the sexiest fields in science to go work on.

That’s kind of a low bar, the sexiest fields in science.

Well, it depends on who you’re appealing to. That’s pretty sexy talk for some people.

And so you may be right, we going to reframe it to get everybody excited. There are whole generations of young people who want to make an impact. And I know very few industries, very few fields, where you can make an impact like climate science, like oceanic science, like environmental sciences and like materials sciences and robotics.

Just as you gravitate around the Large Hadron Collider (a particle accelerator in Europe), just as you gravitate around the Square Kilometer Array (a powerful radio telescope planned in Australia), you gravitate around instruments because that’s where you do your science.

And we have to give Oregon State an amazing instrument so that researchers could do their science and students will go.

You’re the only chip CEO founder still standing in the business. You started in Nvidia in the ‘90s. How do you see this donation? How do you see this fitting into the legacy that you’ve forged over the past, 30, 35 years?

This is surely one of the great capstones of our career, to be able to make a contribution back to Oregon State. And, also, for the impact that we have an opportunity to make.

The size of the company, the success of the company, we’re very proud of. The impact of the company and its reach around the world and in industries that we serve, we can be immensely proud. And now we’re making an investment into the next generation.

The impact could go on for several generations. So we’re proud of that. And it’s also the place where we met, probably one block away from where we met.

Is there one question that you think: This is the sort of thing we might be able to answer out of this kind of work?

Climate science is divisive today. The reason for that is because people see a different future.

Some people think that climate science and climate change is a real problem. Some people don’t. People also see the solutions differently.

We need a time machine. We need a simulation. We need a method to predict the impact of climate science and its magnitude of impact in different regions around.

We can do this. It’s within the capabilities of our technology, within the capabilities of our time, to simulate, to predict the impact of climate change in different regions around the world so that we can answer the question, What does climate change mean to me?

You need to be able to answer these questions. ‘What does it mean to you,’ so that it unites people instead of divides people.

We also need a simulator that allows us to simulate scenarios so that we can predict the impact of the mitigation strategies that we have. Which one do we use first?

Every single mitigation strategy has side effects. We need to be able to simulate its impact as well as its side effects to understand the net benefit to mitigating climate change.

We need amazing scientists, climate scientists, and (to) give them the right instruments, the right tools, a time machine, so that they can go into the future and explain and bring back the answer to us.

I’m very optimistic that they’ll be able to do it.

Is there anything particularly exciting to you about the technology behind (this supercomputer)?

Every part of the computer doesn’t exist today. That’s exciting. In order to build this computer, every aspect of it had to be invented. The CPU had to be invented. The GPU processor had to be invented. The networking had to be invented.

The software and the algorithms, the artificial intelligence algorithms, all has to be invented. That’s kind of cool.

It’s cool for climate scientists, it’s cool for computer scientists, it’s going to be great for the students. It’s going to be great for Oregon and hopefully it’s going to be great for generations.

— Mike Rogoway | mrogoway@oregonian.com | 503-294-7699 | Twitter: @rogoway |

Our journalism needs your support. Please become a subscriber today at OregonLive.com/subscribe.