LongCut logo

Nvidia CEO Jensen Huang on AI, Musk and Trump

By Bloomberg Television

Summary

## Key takeaways - **$45B Q2 Sales Despite $8B China Loss**: Fiscal second quarter sales outlook is $45 billion plus or minus 2%, with $8 billion lost revenue from China made up by strong demand for Blackwell and reasoning inference. [00:04], [01:52] - **Reasoning Inference Demand Explodes**: Demand for reasoning inference is off the charts due to popularity of AI services like ChatGPT and Gemini, making inferencing a giant workload with apps built on top doing incredibly well. [00:26], [00:58] - **China: 2nd Largest AI Market**: China is the second largest AI market and home to the world's largest population of AI researchers; Nvidia wants all world's AI researchers building on American stacks. [01:55], [02:10] - **Huawei Matches H200, Scales Bigger**: Huawei's technology is comparable to H200 and their Cloud Matrix scales larger than Grace Blackwell; they are a formidable company moving fast. [04:18], [04:32] - **Praise Trump's Tariffs & AI Policy**: Trump's tariffs are visionary to re-industrialize America and onshore manufacturing; rescinding AI diffusion rule accelerates American stacks worldwide. [06:56], [07:36] - **Optimus: Next Multitrillion Industry**: Tesla's Optimus is world-class and revolutionary, the first robot to achieve high volume and tech scale; likely the next multitrillion dollar industry. [10:48], [11:13]

Topics Covered

  • Blackwell Offsets China Revenue Loss
  • China's AI Researchers Build American Stack
  • Huawei Matches H200, Scales Larger
  • Tariffs Rescind AI Diffusion Rule
  • Optimus Sparks Multitrillion Robot Industry

Full Transcript

You went into great detail about what's happening, particularly in China.

I just wanted to go over something that I felt wasn't asked, which is fiscal second quarter sales, $45 billion plus or -2%.

$8 billion of lost revenue opportunity specifically relating to China and 20.

It seems as if you made up somewhere some demand for a different product or from a different geography or from a different corner of the market.

What was that? Well, we got a whole bunch of engines firing right now. The biggest one, of course, is the reasoning inference.

The demand is just off the charts. You see, the popularity of all these air services now charge up to Gemini, you know, so on, so forth.

I mean, they're just doing incredibly well across the board and all of the apps that they serve out and all the Genting air services that are built on top of them. They're all doing incredibly well.

I think that there's just a giant breakthrough and an ACE capability and inferencing has just become a giant workload.

Second, now that people realize that Black Wall is just a homerun and Building 72 is a home run architecture, we designed it to be a thinking machine, a reasoning A.I. system.

And I think people now the confluence of the the the the breakthrough in reasoning A.I. and the availability that, you know, the

reasoning A.I. and the availability that, you know, the the now the emergence of of Grace Blackwell and being like 72 perfect timing. I think that that's at the core, a big

timing. I think that that's at the core, a big part of it. In the second part of it is that our supply chain is growing and we're really ramping it up and they're doing fantastically for us. And so all of these things are all coming together. And so it is fair to say that some of

coming together. And so it is fair to say that some of that additional supply on Blackwell and the demand for Blackwell kind of made up for the opportunity lost in China, at least in the outlook for this current period. Yeah, I guess so.

period. Yeah, I guess so.

But, you know, you just can't you can't underestimate the importance of the China market. This is the second largest air market.

China market. This is the second largest air market.

This is the home of the world's largest population of air researchers.

And we want all of the world's air researchers and all of the world's developers to be building on American stocks.

And so, you know, irrespective of of the near term revenue success that we have, we can't ignore the fact that the Chinese market is very important.

You explained again in some detail that at least in the hopper architecture, you have engineered down to the lowest spec possible.

It's not possible to do anything different with Hopper for the Chinese market, but in your consideration for a different architecture or a chip for the Chinese market, is that what we're talking about, a new design or a new class? And have you made that proposal to the

class? And have you made that proposal to the administration just designing from the ground up a new chip?

We're still thinking through that. They're just

the limitations are quite, quite stringent, quite limited, if you will.

H-20 is is, you know, as far down as we could take a hopper, we don't know how to make it even even less.

And so that's really the limit. But but so there aren't there aren't many you know, the limitations are quite stringent.

So we we have to really think through what whatever we make ultimately has to add value to the market. And so it's it's a really tightrope because because the Chinese competitors have evolved and advanced greatly over the last year. Like everybody else, they're, you know, doubling, quadrupling their capabilities every year.

And the volume is increasing substantially.

And remember, these are data center chips.

They don't have to be small. They could be quite large.

And and and, you know, without American technology, the availability of Chinese technology will fill the market. And so, you know, whatever we offer has to at least at least be competitive and has to add value to the market.

Jensen just while we have an accelerator or a GPU that is performance, is H20 or is performance of other classes of GPU that you make?

Huawei's technology based on based on our our best understanding at the moment. And we have a lot of ground truth there

moment. And we have a lot of ground truth there is probably comparable to an H 200. And so they're there.

They've been moving quite fast. And they've also offered this new system called Cloud matrix, that is that is scales up to even a larger system than our latest generation, Grace Blackwell. And so so their Huawei, as you know, is a formidable technology company. And they're not sitting still.

And they they look for ways to look for ways to compete.

And they're quite formidable. With that in mind, you've kind of explained the landscape now Nvidia's ability to operate in China.

I mean, you and I this year alone have already discussed the idea that there are 50% of the world's A.I. researchers in or from China.

But are you having a sense that the vacuum created that those big names I think about Tencent or Alibaba or Baidu that were buyers of age 20, that they've already pivoted and turned to the offering from far away because of the policy that is in place. Yeah, they have no choice but to, you

know, one of our one of the one of the challenges of the changing regulations is the ability for markets to trust the media and and ultimately American platforms. And so so it's prudent, I think, for the the Chinese customers to make sure that they they develop their stack on Huawei.

And because it's hard to rely on American technology at this point.

And so that's one of the that's one of the unfortunate, unfortunate parts of of changing policies. But anyhow, you know, I have every

changing policies. But anyhow, you know, I have every confidence that that if able to compete, American companies will compete.

This is in it to to to to write off American technology companies is is not smart This is this is the home of of of some of the brightest computer scientists in the world.

American companies are incredibly competitive.

We just have to have the confidence to go compete.

And if we have the confidence to compete, we will win.

You said during the earnings call that you trust President Trump and that the president has a vision and a plan. Can I ask if you've talked to him about that plan and if it includes coordinating with India on policy adjustment that relates to changes in technology export controls?

Well, obviously, I don't know all of his ideas, but let me tell you about two that are incredible. The first one is utterly visionary, the idea of tariffs being a pillar of goal, of a bold vision to re industrialize, to onshore manufacturing and motivate the world to invest in the United States.

And it's just an incredible vision. I think this is going to be a transformative idea for the next century for us.

We're all in on the idea where we're setting up plants and encouraging our partners from around the world to invest in our states.

And we have a lot of stuff going on. And so I'm very excited about that.

The second the second major idea is to rescind the air diffusion rule, recognizing that this isn't about limiting American technology, but this is about accelerating American stocks around the world to make sure that before it's too late, that the world builds on on American stocks during this extraordinary time, the era. And so these two these two initiatives

are completely visionary, and it's going to be transformative for America.

Jensen, in the time that you and I have been on air having this conversation, some news is broken from US Secretary of State Rubio, who has said that the US will begin revoking some Chinese student visas.

What I wanted to ask you is that with US government limits on foreign student visas, how does that impact a company like NVIDIA?

I think about the size, but also composition of your engineering talent here in California and elsewhere in the United States.

I believe the administration still feels very strongly about about the incredible importance of immigration.

Look, I'm an immigrant. I know many immigrants that came in can't United States to to build a great life.

And many of us many of us have contributed greatly to the technology industry and in the United States. I believe that that's going to have to continue. Remember, remember, you know, people

continue. Remember, remember, you know, people from all over the world want to come to United States.

This is such an extraordinary country with such incredible opportunities.

We want the brightest to come here. We don't want we don't want others to you know, we don't want everybody to be able to come here.

And there should be there should be rules.

And and but nonetheless, for the ones that really can make a contribution, we want to make a difference and we want to make it possible for them to come here and and bring their great ideas, bring their great intellect and help us build a great America. And so I think I think the administration is all in on that. And I don't think anything that they've said changes that. Jensen.

I surveyed our Bloomberg Technology audience around the world.

Four questions for you. And I think the most common question is understanding who Nvidia's customers are away from the hyperscalers.

But actually many questions about Elon Inc and whether Tesla and A.I.

in aggregate might actually be one of the biggest customers you have.

You think about not just the data center chips, but this chip specific for Optimus or in the cores, the omniverse component.

Could you speak a little about that? I we do.

We do a lot of business with Tesla. And I say I in Lonza Airlines that, as you know, we're just an extraordinary engineer and I love working with him.

We've built some amazing computers together.

We're going to build many more computers together.

And and the work that he's doing in Grok his self-driving car, his Optimus. These are these are all every single one

his Optimus. These are these are all every single one of them. World class, every single one of them

of them. World class, every single one of them revolutionary. Every single one of them are going to be

revolutionary. Every single one of them are going to be gigantic opportunities. And and we're delighted.

gigantic opportunities. And and we're delighted.

I'm delighted to be working with them on that.

And so I, I think the the the Optimus Opportunity is just right around the corner. It's very likely that human robots are are going to be robots that we can deploy into the world relatively easily.

And this is the first robot that really has a chance to achieve the high volume and technology scale necessary to advance technology.

And and so I think this is likely to be the next multitrillion dollar industry.

I'm very excited for him. Jensen Let's end by talking about Europe. You alluded to a trip you're taking next

Europe. You alluded to a trip you're taking next week. It sounds as if the infrastructure build

week. It sounds as if the infrastructure build out in Europe is coming. Where will you be visiting and who will you be speaking with? Well, I'm going to leave the Who as a surprise for all of you. But.

But I'll be seeing lots of heads of states and I'll be in France and UK and Germany and Belgium.

And and I think that that the the it's very clear now that every every country recognizes that artificial intelligence, like electricity, like Internet, like communications is part of a national infrastructure.

No, no society could do without intelligence, as you know.

And and there's an awakening that every country has to have, has to take some initiative to ensure that that their country and their society has access to artificial intelligence. And so we're we're really delighted to

artificial intelligence. And so we're we're really delighted to be able to to work with, you know, work with the European countries to to bring infrastructure development, work with them to build a factory. There's an umpteen number of air factory

factory. There's an umpteen number of air factory projects in discussion and development. And and so I'm really excited to make this trip. We're going to be all over Europe.

this trip. We're going to be all over Europe.

Loading...

Loading video analysis...