WEBVTT - Nvidia CEO Jensen Huang Talks Earnings, AI Growth Outlook, China Market

0:00:02.520 --> 0:00:07.400
<v Speaker 1>Bloomberg Audio Studios, podcasts, radio news.

0:00:07.840 --> 0:00:10.560
<v Speaker 2>Let's bring in and video CEO Jensen Wang following the

0:00:10.560 --> 0:00:14.240
<v Speaker 2>company's earnings and analysts call Jensen, thank you for your

0:00:14.280 --> 0:00:18.520
<v Speaker 2>time here on Bloomberg Technology. You went into great detail

0:00:19.040 --> 0:00:22.239
<v Speaker 2>about what's happening, particularly in China. I just wanted to

0:00:22.239 --> 0:00:24.479
<v Speaker 2>go over something that I felt wasn't asked, which is

0:00:25.239 --> 0:00:28.639
<v Speaker 2>fiscal second quarter sales forty five billion dollars plus or

0:00:28.680 --> 0:00:33.240
<v Speaker 2>minus two percent eight billion dollars of lost revenue opportunity,

0:00:33.520 --> 0:00:36.960
<v Speaker 2>specifically relating to China in eight twenty. It seems as

0:00:37.000 --> 0:00:40.640
<v Speaker 2>if you made up somewhere some demand for a different product,

0:00:40.760 --> 0:00:43.240
<v Speaker 2>or from a different geography, or from a different corner

0:00:43.240 --> 0:00:43.959
<v Speaker 2>of the market.

0:00:44.240 --> 0:00:47.479
<v Speaker 1>What was that, Well, we got a.

0:00:47.440 --> 0:00:50.880
<v Speaker 3>Whole bunch of engines firing right now. The biggest one,

0:00:50.880 --> 0:00:55.160
<v Speaker 3>of course, is the reasoning AI inference. The demand is

0:00:55.240 --> 0:01:00.320
<v Speaker 3>just after the charts, you see the popularity of all

0:01:00.360 --> 0:01:04.320
<v Speaker 3>these AI services now Chat, GBT, Gemini, you know, so

0:01:04.400 --> 0:01:06.360
<v Speaker 3>on and so forth, GROC. I mean, they're just doing

0:01:06.400 --> 0:01:09.679
<v Speaker 3>incredibly well across the board, and all of the APIs

0:01:09.720 --> 0:01:12.679
<v Speaker 3>that they serve out and all the agentic AI services

0:01:12.680 --> 0:01:15.160
<v Speaker 3>that are built on top of them, they're all doing

0:01:15.240 --> 0:01:18.480
<v Speaker 3>incredibly well. I think that there's just a giant breakthrough

0:01:19.040 --> 0:01:24.840
<v Speaker 3>in AI's capability, and inferencing has just become a giant workload. Second,

0:01:25.640 --> 0:01:28.120
<v Speaker 3>people realize that Blackwell is just a home run envey

0:01:28.160 --> 0:01:31.000
<v Speaker 3>Link seventy two is a home run architecture. We designed

0:01:31.040 --> 0:01:34.120
<v Speaker 3>it to be a thinking machine, a reasoning AI system.

0:01:34.440 --> 0:01:38.720
<v Speaker 3>And I think people now the confluence of the breakthrough

0:01:38.760 --> 0:01:42.600
<v Speaker 3>and reasoning AI and the availability you know now the

0:01:42.640 --> 0:01:46.240
<v Speaker 3>emergence of Grace Blackwell and VEALINGK seventy two perfect timing.

0:01:46.520 --> 0:01:49.800
<v Speaker 1>I think that that's at the core a big part

0:01:49.840 --> 0:01:50.000
<v Speaker 1>of it.

0:01:50.040 --> 0:01:51.880
<v Speaker 3>And the second part of it is that our supply

0:01:51.960 --> 0:01:54.320
<v Speaker 3>chain is growing and we're really ramping it up and

0:01:54.360 --> 0:01:57.560
<v Speaker 3>they're doing fantastically for us, and so all of these

0:01:57.560 --> 0:01:58.720
<v Speaker 3>things are all coming together.

0:01:59.560 --> 0:02:01.440
<v Speaker 2>And so it is fair to say that some of

0:02:01.480 --> 0:02:04.600
<v Speaker 2>that additional supply on Blackwell and the demand for black

0:02:04.640 --> 0:02:08.239
<v Speaker 2>Bell kind of made up for the opportunity lost in China,

0:02:08.280 --> 0:02:10.560
<v Speaker 2>at least in the outlook for this current period.

0:02:12.280 --> 0:02:13.040
<v Speaker 1>Yeah, I guess so.

0:02:13.280 --> 0:02:17.200
<v Speaker 3>But you know, you just can't you can't underestimate the

0:02:17.240 --> 0:02:21.560
<v Speaker 3>importance of the China market. This is the second largest

0:02:21.639 --> 0:02:25.800
<v Speaker 3>AI market, This is the home of the world's largest

0:02:25.800 --> 0:02:29.880
<v Speaker 3>population of AI researchers, and we want all of the

0:02:29.919 --> 0:02:32.680
<v Speaker 3>world's AI researchers and all of the world's developers to

0:02:32.680 --> 0:02:36.639
<v Speaker 3>be building on American stacks. And so, you know, irrespective

0:02:36.680 --> 0:02:41.080
<v Speaker 3>of the nearer term revenue success that we have, we

0:02:41.120 --> 0:02:43.800
<v Speaker 3>can't ignore the fact that the Chinese market is very important.

0:02:44.880 --> 0:02:48.480
<v Speaker 2>You explained again in some detail that at least in

0:02:48.520 --> 0:02:52.239
<v Speaker 2>the Hopper architecture you have engineered down to the lowest

0:02:52.280 --> 0:02:56.320
<v Speaker 2>spec possible. It's not possible to do anything different with

0:02:56.440 --> 0:03:00.480
<v Speaker 2>Hopper for the Chinese market. But in your conceit iteration

0:03:00.760 --> 0:03:04.000
<v Speaker 2>for a different architecture or a chip for the Chinese market,

0:03:04.440 --> 0:03:06.880
<v Speaker 2>is that what we're talking about? A new design or

0:03:06.919 --> 0:03:10.400
<v Speaker 2>new class? And have you made that proposal to the administration?

0:03:10.720 --> 0:03:12.880
<v Speaker 2>Just designing from the ground up and new chip.

0:03:13.480 --> 0:03:15.000
<v Speaker 1>We're still thinking through that.

0:03:15.840 --> 0:03:21.440
<v Speaker 3>They're just the limitations are quite quite stringent, quite limited,

0:03:21.480 --> 0:03:22.000
<v Speaker 3>if you will.

0:03:22.919 --> 0:03:23.760
<v Speaker 1>H twenty is.

0:03:25.760 --> 0:03:28.160
<v Speaker 3>You know, as far down as we could take a Hopper,

0:03:28.639 --> 0:03:30.919
<v Speaker 3>we don't know how to make it even even less,

0:03:31.520 --> 0:03:35.680
<v Speaker 3>and so that's really the limit. But but so there

0:03:35.680 --> 0:03:39.080
<v Speaker 3>aren't there aren't you know, the limitations are quite stringent.

0:03:39.200 --> 0:03:40.920
<v Speaker 1>So we have we have to really think through it.

0:03:41.560 --> 0:03:44.680
<v Speaker 3>Whatever we make ultimately has to add value to the market.

0:03:44.960 --> 0:03:48.560
<v Speaker 3>And so it's it's a really tight rope because because

0:03:49.320 --> 0:03:53.360
<v Speaker 3>the Chinese competitors have evolved and advanced greatly over the

0:03:53.400 --> 0:03:57.600
<v Speaker 3>last year like everybody else, they're you know, doubling, quadrupling

0:03:57.600 --> 0:04:01.760
<v Speaker 3>their capabilities every year, and the vlume is increasing substantially.

0:04:01.800 --> 0:04:03.960
<v Speaker 3>And remember these are data center chips. They don't have

0:04:04.040 --> 0:04:07.920
<v Speaker 3>to be small, they could be quite large and and

0:04:07.600 --> 0:04:13.280
<v Speaker 3>and you know they without American technology, UH, the availability

0:04:13.280 --> 0:04:16.960
<v Speaker 3>of Chinese technology will fail the market. And so you know,

0:04:16.960 --> 0:04:19.680
<v Speaker 3>we whatever we offer has to at least at least

0:04:19.760 --> 0:04:22.159
<v Speaker 3>be competitive and has to add value to the market.

0:04:23.440 --> 0:04:27.440
<v Speaker 2>Tensen, Does Huawei have an AI accelerator or a GPU

0:04:27.600 --> 0:04:31.080
<v Speaker 2>that is performance is H twenty or is performance of

0:04:31.120 --> 0:04:33.200
<v Speaker 2>other classes of GPU that you make?

0:04:35.120 --> 0:04:39.039
<v Speaker 3>Uh, Huawei's technology based on based on our our best

0:04:39.200 --> 0:04:41.560
<v Speaker 3>understanding at the moment, UH, and we have a lot

0:04:41.560 --> 0:04:45.719
<v Speaker 3>of ground truth there is probably comparable to an age

0:04:45.800 --> 0:04:50.520
<v Speaker 3>two hundred and so they're they're they've been moving quite

0:04:50.600 --> 0:04:55.320
<v Speaker 3>fast and and they've also offered UH this news system called.

0:04:55.120 --> 0:04:59.520
<v Speaker 1>Cloud Matrix UH that is that is scales.

0:04:59.160 --> 0:05:02.320
<v Speaker 3>Up to even a larger system than our latest generation

0:05:02.440 --> 0:05:06.360
<v Speaker 3>Grace Blackwell and so there. Huawei, as you know, is

0:05:06.400 --> 0:05:09.359
<v Speaker 3>a formidable technology company and they're not sitting still and

0:05:09.400 --> 0:05:12.400
<v Speaker 3>they look for ways to look for ways to compete,

0:05:12.880 --> 0:05:14.920
<v Speaker 3>and they're quite formidable.

0:05:17.000 --> 0:05:19.440
<v Speaker 2>With that in mind, you've kind of explained the landscape

0:05:19.480 --> 0:05:22.200
<v Speaker 2>now in Vidia's ability to operate in China. I mean,

0:05:22.240 --> 0:05:24.320
<v Speaker 2>you and I this year alone have already discussed the

0:05:24.360 --> 0:05:28.080
<v Speaker 2>idea that they're fifty percent of the world's AI researchers

0:05:28.520 --> 0:05:31.239
<v Speaker 2>in or from China. But are you having a sense

0:05:31.279 --> 0:05:34.640
<v Speaker 2>of the vacuum created that those big names I think

0:05:34.680 --> 0:05:36.880
<v Speaker 2>about ten Cent or Ali Baba or by Do that

0:05:36.920 --> 0:05:40.200
<v Speaker 2>were buyers of age twenty that they've already pivoted and

0:05:40.279 --> 0:05:44.839
<v Speaker 2>turned to the offering from Huawei because of the policy

0:05:44.880 --> 0:05:45.599
<v Speaker 2>that is in place.

0:05:47.279 --> 0:05:49.880
<v Speaker 3>Yeah, they have no choice, but to you know, of

0:05:49.960 --> 0:05:52.919
<v Speaker 3>our one of the one of the challenges of the

0:05:53.040 --> 0:05:57.480
<v Speaker 3>changing regulations is the ability for markets to trust the

0:05:57.600 --> 0:06:03.080
<v Speaker 3>Nvidia and ultimately emerge CAN platforms and so so, Uh,

0:06:03.400 --> 0:06:08.080
<v Speaker 3>it's prudent I think for the Chinese customers uh to

0:06:08.240 --> 0:06:12.000
<v Speaker 3>make sure that they they develop their stack on Huawei

0:06:12.600 --> 0:06:16.480
<v Speaker 3>and because it's hard to rely on American technology at

0:06:16.480 --> 0:06:18.560
<v Speaker 3>this point, and so that's one of the that's one

0:06:18.560 --> 0:06:22.680
<v Speaker 3>of the the unfortunate, unfortunate parts of of changing policies.

0:06:24.320 --> 0:06:28.000
<v Speaker 3>But anyhow, you know, I have every confidence that that

0:06:28.200 --> 0:06:32.440
<v Speaker 3>if able to compete, American companies will compete. This is

0:06:32.800 --> 0:06:36.960
<v Speaker 3>in a to to uh uh, to write off American

0:06:37.000 --> 0:06:39.800
<v Speaker 3>technology companies is not smart.

0:06:39.800 --> 0:06:43.040
<v Speaker 1>This is this is the home of of of.

0:06:42.920 --> 0:06:46.560
<v Speaker 3>Some of the brightest computer scientists in the world. American

0:06:46.600 --> 0:06:49.880
<v Speaker 3>companies are incredibly competitive. We just have to have the

0:06:49.880 --> 0:06:52.240
<v Speaker 3>confidence to go compete, and if we have the confidence

0:06:52.240 --> 0:06:53.159
<v Speaker 3>to compete, we will win.

0:06:54.960 --> 0:06:58.040
<v Speaker 2>You said during the ending school that you trust President

0:06:58.080 --> 0:07:01.599
<v Speaker 2>Trump and that the president has an vision and a plan.

0:07:02.080 --> 0:07:04.120
<v Speaker 2>Could I ask if you've talked to him about that

0:07:04.279 --> 0:07:08.560
<v Speaker 2>plan and if it includes coordinating with Nvidia on policy

0:07:08.560 --> 0:07:12.720
<v Speaker 2>adjustment that relates to changes in technology export controls.

0:07:13.960 --> 0:07:17.120
<v Speaker 3>Well, obviously I don't know all of his ideas, but

0:07:17.240 --> 0:07:20.000
<v Speaker 3>let me tell you about two that are incredible. The

0:07:20.040 --> 0:07:24.600
<v Speaker 3>first one is utterly visionary, the idea of tariffs being

0:07:24.640 --> 0:07:30.080
<v Speaker 3>a pillar of a bold vision to reindustrialize to onshore

0:07:30.160 --> 0:07:35.400
<v Speaker 3>manufacturing and motivate the world to invest in the United States.

0:07:35.840 --> 0:07:38.880
<v Speaker 3>And it's just an incredible vision. I think this is

0:07:38.920 --> 0:07:41.880
<v Speaker 3>going to be a transformative idea for the next century

0:07:41.880 --> 0:07:44.880
<v Speaker 3>for us. We're all in on the idea where we're

0:07:45.120 --> 0:07:48.280
<v Speaker 3>setting up plants and encouraging our partners from around the

0:07:48.320 --> 0:07:50.280
<v Speaker 3>world to invest the United States, and we have a

0:07:50.280 --> 0:07:52.640
<v Speaker 3>lot of stuff going on, and so I'm very excited

0:07:52.640 --> 0:07:55.800
<v Speaker 3>about that. The second, the second major idea is to

0:07:55.880 --> 0:07:59.800
<v Speaker 3>we send the AI Diffusion Rule, recognizing that this is

0:07:59.800 --> 0:08:04.600
<v Speaker 3>an about limiting American technology, but this is about accelerating

0:08:04.800 --> 0:08:08.320
<v Speaker 3>American stacks around the world to make sure that before

0:08:08.360 --> 0:08:11.960
<v Speaker 3>it's too late, that the world builds on American stacks

0:08:12.360 --> 0:08:16.960
<v Speaker 3>during this extraordinary time the era. And so these two

0:08:17.040 --> 0:08:20.480
<v Speaker 3>initiatives are completely visionary and it's going to be transformative

0:08:20.480 --> 0:08:21.040
<v Speaker 3>for America.

0:08:22.480 --> 0:08:24.200
<v Speaker 2>Jensen. In the time that you and I have been

0:08:24.320 --> 0:08:26.960
<v Speaker 2>on air having this conversation, some news is broken from

0:08:27.120 --> 0:08:30.040
<v Speaker 2>US Secretary of State Rubio who has said that the

0:08:30.120 --> 0:08:34.440
<v Speaker 2>US will begin revoking some Chinese student visas. What I

0:08:34.480 --> 0:08:37.520
<v Speaker 2>wanted to ask you is that with US government limits

0:08:37.600 --> 0:08:40.920
<v Speaker 2>on foreign student visas, how does that impact a company

0:08:41.080 --> 0:08:44.800
<v Speaker 2>like Nvidia. I think about the size but also composition

0:08:44.880 --> 0:08:48.199
<v Speaker 2>of your engineering talent here in California and elsewhere in

0:08:48.240 --> 0:08:49.160
<v Speaker 2>the United States.

0:08:52.400 --> 0:08:56.440
<v Speaker 3>I believe the administration still feels very strongly about about

0:08:57.640 --> 0:08:59.480
<v Speaker 3>the incredible importance of immigration.

0:09:00.160 --> 0:09:01.360
<v Speaker 1>Look, I'm an immigrant.

0:09:01.840 --> 0:09:04.520
<v Speaker 3>Uh. I know many immigrants that that came and came

0:09:04.559 --> 0:09:07.120
<v Speaker 3>to United States to to build a great life. And

0:09:07.240 --> 0:09:10.840
<v Speaker 3>many of us, many of us have contributed greatly to

0:09:10.920 --> 0:09:13.719
<v Speaker 3>the technology industry and and uh in.

0:09:13.760 --> 0:09:16.560
<v Speaker 1>The United States. I believe that that's going to have

0:09:16.600 --> 0:09:18.120
<v Speaker 1>to continue. Uh.

0:09:18.679 --> 0:09:22.040
<v Speaker 3>Remember, remember, you know, people from all over the world

0:09:22.120 --> 0:09:24.640
<v Speaker 3>want to come to United States. This is just such

0:09:24.640 --> 0:09:27.679
<v Speaker 3>an extraordinary country with such incredible opportunities.

0:09:28.080 --> 0:09:29.600
<v Speaker 1>We want the brightest to come here.

0:09:29.800 --> 0:09:32.160
<v Speaker 3>We don't want we don't want others to you know,

0:09:32.200 --> 0:09:34.360
<v Speaker 3>we we don't want everybody to be able to come here.

0:09:34.640 --> 0:09:37.160
<v Speaker 3>And there should be there should be rules and and

0:09:37.960 --> 0:09:40.559
<v Speaker 3>but nonetheless for the ones that that really can make

0:09:40.559 --> 0:09:43.079
<v Speaker 3>a contribution, want to make a difference, and we want

0:09:43.120 --> 0:09:45.200
<v Speaker 3>to make it possible for them to come here and

0:09:45.200 --> 0:09:47.880
<v Speaker 3>and bring their great ideas, bring their great intellect and

0:09:48.120 --> 0:09:50.320
<v Speaker 3>help us build in a great America. And so I

0:09:50.360 --> 0:09:52.439
<v Speaker 3>think I think the administration is all in on that,

0:09:52.520 --> 0:09:55.200
<v Speaker 3>and I don't think anything that they've said changes that.

0:09:56.400 --> 0:09:59.520
<v Speaker 2>Jensen I surveyed our Bloomberg Technology audience around the world

0:09:59.600 --> 0:10:02.160
<v Speaker 2>for questions for you, and I think the most common

0:10:02.240 --> 0:10:07.240
<v Speaker 2>question is understanding who Nvidia's customers are away from the hyperscalers,

0:10:07.520 --> 0:10:11.720
<v Speaker 2>but actually many questions about Elon Inc. And whether Tesla

0:10:11.880 --> 0:10:15.280
<v Speaker 2>and Xai and Aggregate might actually be one of the

0:10:15.280 --> 0:10:17.480
<v Speaker 2>biggest customers you have. You think about not just the

0:10:17.559 --> 0:10:22.400
<v Speaker 2>data center chips, but this chip specific for optimists or

0:10:22.400 --> 0:10:25.000
<v Speaker 2>in the cars, the omniverse component.

0:10:25.040 --> 0:10:29.040
<v Speaker 1>Could you speak a little about that. We do.

0:10:29.320 --> 0:10:34.840
<v Speaker 3>We do a lot of business with Tesla and Xai Elons. Elons,

0:10:35.200 --> 0:10:38.440
<v Speaker 3>as you know, is just an extraordinary engineer and I

0:10:38.480 --> 0:10:41.640
<v Speaker 3>love working with him. We've built some amazing computers together.

0:10:41.800 --> 0:10:46.200
<v Speaker 3>We're going to build many more computers together. And the

0:10:46.200 --> 0:10:51.080
<v Speaker 3>work that he's doing in groc his self driving car,

0:10:52.480 --> 0:10:55.920
<v Speaker 3>his optimists, these are these are all, every single one

0:10:55.920 --> 0:10:58.880
<v Speaker 3>of them world class, every single one of them revolutionary,

0:10:58.920 --> 0:11:01.840
<v Speaker 3>every single one of them are going to be gigantic opportunities,

0:11:01.840 --> 0:11:05.920
<v Speaker 3>and we're delighted. I'm delighted to be working with him

0:11:05.920 --> 0:11:06.160
<v Speaker 3>on that.

0:11:06.320 --> 0:11:08.080
<v Speaker 1>And so I think the.

0:11:10.240 --> 0:11:13.839
<v Speaker 3>The optimist opportunities just right around the corner. It's very

0:11:13.960 --> 0:11:17.480
<v Speaker 3>likely that human robots are are going to be robots

0:11:17.480 --> 0:11:22.360
<v Speaker 3>that we can deploy into the world relatively easily. And

0:11:22.720 --> 0:11:25.120
<v Speaker 3>this is the first robot that really has a chance

0:11:25.559 --> 0:11:29.760
<v Speaker 3>to achieve the high volume and technology scale necessary to

0:11:29.800 --> 0:11:30.760
<v Speaker 3>advance technology.

0:11:30.800 --> 0:11:32.560
<v Speaker 1>And so I think this.

0:11:32.440 --> 0:11:35.520
<v Speaker 3>Is likely to be the next multi trillion dollar industry.

0:11:35.920 --> 0:11:36.959
<v Speaker 1>I'm very excited for him.

0:11:37.960 --> 0:11:41.000
<v Speaker 2>Tensen. Let's end by talking about Europe. You alluded to

0:11:41.040 --> 0:11:44.080
<v Speaker 2>a trip you're taking next week. It sounds as if

0:11:44.120 --> 0:11:48.040
<v Speaker 2>the infrastructure build out in Europe is coming. Where will

0:11:48.080 --> 0:11:49.840
<v Speaker 2>you be visiting and who we be speaking with?

0:11:51.200 --> 0:11:54.040
<v Speaker 3>Well, I'm going to leave the who as a surprise

0:11:54.120 --> 0:11:56.920
<v Speaker 3>for all of you, but I'll be seeing lots of

0:11:57.440 --> 0:12:00.960
<v Speaker 3>heads of states, and I'll be in France, in UK

0:12:01.240 --> 0:12:06.920
<v Speaker 3>and Germany, UH and UH Belgium, and and I think

0:12:06.920 --> 0:12:11.200
<v Speaker 3>that that the the It's very clear now that every

0:12:11.280 --> 0:12:16.720
<v Speaker 3>under every country recognizes that artificial intelligence, like electricity, like internet,

0:12:16.800 --> 0:12:20.800
<v Speaker 3>like communications, is part of a national infrastructure. No, no

0:12:21.000 --> 0:12:24.640
<v Speaker 3>society could do without intelligence, as you know, And and

0:12:24.920 --> 0:12:28.360
<v Speaker 3>there's an awakening that every country has to has to

0:12:28.400 --> 0:12:32.360
<v Speaker 3>take some initiative to ensure that that their country and

0:12:32.400 --> 0:12:36.400
<v Speaker 3>their society has access to artificial intelligence. And so we're

0:12:36.440 --> 0:12:39.000
<v Speaker 3>we're really delighted to be able to to work with

0:12:39.160 --> 0:12:43.000
<v Speaker 3>uh uh uh uh. You know, work with the European

0:12:43.520 --> 0:12:47.679
<v Speaker 3>UH countries to to bring AI infrastructure to them and

0:12:48.200 --> 0:12:50.880
<v Speaker 3>work with them to build AI factories. There's an UN

0:12:51.000 --> 0:12:55.240
<v Speaker 3>team number of AI factory projects in discussion and development

0:12:55.320 --> 0:12:58.800
<v Speaker 3>and and so so I'm really excited to make this trip.

0:12:58.840 --> 0:12:59.960
<v Speaker 1>We're going to be all over Europe.

0:13:00.040 --> 0:13:00.240
<v Speaker 3>MH.

0:13:01.600 --> 0:13:05.559
<v Speaker 2>Convidia CEO Jansen Wang, thank you, thank you, good to

0:13:05.600 --> 0:13:05.920
<v Speaker 2>see you.