WEBVTT - Bloomberg Tech Special: Nvidia CEO Jensen Huang on Upbeat Forecast

0:00:02.480 --> 0:00:07.000
<v Speaker 1>Bloomberg Audio Studios, Podcasts, Radio News.

0:00:07.720 --> 0:00:10.280
<v Speaker 2>We'd like to welcome in video CEO Jensen Wang to

0:00:10.360 --> 0:00:14.160
<v Speaker 2>the program, and Jensen, it's been an astonishingly busy day

0:00:14.200 --> 0:00:17.560
<v Speaker 2>for you in Washington, DC. So I'm grateful for your time.

0:00:18.400 --> 0:00:22.400
<v Speaker 2>You're sold out of Blackwell, that's what you said, but

0:00:22.520 --> 0:00:25.840
<v Speaker 2>also that five hundred billion dollar forecast, which is Blackwell

0:00:25.960 --> 0:00:30.160
<v Speaker 2>Rubin has room to grow. How do those fit together?

0:00:34.040 --> 0:00:40.760
<v Speaker 1>I said, sales are after charts for Blackwell and in

0:00:40.880 --> 0:00:44.879
<v Speaker 1>video GPUs in the cloud are sold out. We got

0:00:44.880 --> 0:00:47.879
<v Speaker 1>plenty of Blackwells to sell you. We have lots of

0:00:47.920 --> 0:00:51.920
<v Speaker 1>Blackwells coming, We're making a lot of Blackwells, and we

0:00:51.960 --> 0:00:57.040
<v Speaker 1>have a bunch of Vera Rubens coming and so business

0:00:57.120 --> 0:01:01.000
<v Speaker 1>is very very strong. But we we've planned our supply

0:01:01.080 --> 0:01:03.640
<v Speaker 1>chain incredibly well. We have the largest supply chain in

0:01:03.640 --> 0:01:09.759
<v Speaker 1>the world. Our partners TSMC, our memory partners, s k Heinex, Micron,

0:01:09.840 --> 0:01:13.520
<v Speaker 1>Samsung are doing a fantastic job supporting us. And all

0:01:13.520 --> 0:01:16.680
<v Speaker 1>of our systems partners, fox Con and Quantum and Wistron

0:01:17.200 --> 0:01:20.600
<v Speaker 1>are packaging partners. Everybody's doing a fantastic job supporting us.

0:01:20.680 --> 0:01:23.520
<v Speaker 1>And we've done a good job planning for a very

0:01:23.680 --> 0:01:26.279
<v Speaker 1>very strong year, and we've done a good job planning

0:01:26.280 --> 0:01:30.760
<v Speaker 1>for Vera Rubin, so sales are off the charts and

0:01:30.880 --> 0:01:34.080
<v Speaker 1>video GPUs and the cloud is sold out, but we've

0:01:34.080 --> 0:01:35.600
<v Speaker 1>got a bunch of Blackwells to sell.

0:01:36.800 --> 0:01:39.480
<v Speaker 2>Jensen, what's the road ahead for Vera Rubin. It's one

0:01:39.520 --> 0:01:41.560
<v Speaker 2>of the most common questions we get for you of

0:01:41.640 --> 0:01:44.720
<v Speaker 2>how that ramp will go relative to what we saw

0:01:44.800 --> 0:01:46.880
<v Speaker 2>with the Blackwell generations.

0:01:49.600 --> 0:01:54.640
<v Speaker 1>Well, the silicon for Vera Rubin, seven different chips are

0:01:54.680 --> 0:01:57.320
<v Speaker 1>back in our labs and the bring up is happening

0:01:57.360 --> 0:02:01.480
<v Speaker 1>across engineering teams. Probably a couple of twenty thousand people

0:02:01.480 --> 0:02:05.760
<v Speaker 1>are working on bringing up Vera Rubin. From silicon to systems,

0:02:05.800 --> 0:02:09.880
<v Speaker 1>to software to algorithms. People are working around the clock

0:02:10.400 --> 0:02:14.360
<v Speaker 1>and this bring up is going beautifully. We're on track

0:02:14.480 --> 0:02:18.720
<v Speaker 1>to deliver Vera Rubin about Q three timeframe of next year,

0:02:19.960 --> 0:02:24.560
<v Speaker 1>continuing our once a year cycle. Vera Ruben is already

0:02:25.520 --> 0:02:30.280
<v Speaker 1>assured a huge success. Everybody's incredibly excited about it. Can't

0:02:30.280 --> 0:02:34.280
<v Speaker 1>wait to show everybody. And then one last thing is

0:02:34.320 --> 0:02:39.280
<v Speaker 1>that the Rack architecture, the Rack scale architecture, is completely revolutionary.

0:02:39.919 --> 0:02:43.600
<v Speaker 1>It includes a scale up switch called the mvy link

0:02:44.280 --> 0:02:47.480
<v Speaker 1>mvy Link seventy two. Our fifth generation is the only

0:02:47.560 --> 0:02:50.919
<v Speaker 1>one of its kind in the world. This rack architecture,

0:02:51.360 --> 0:02:55.480
<v Speaker 1>which is incredibly complex, started with Grace Blackwell, then Grace

0:02:55.480 --> 0:03:00.280
<v Speaker 1>Blackwell Ultra. It is transitioned to Grace Blackwell Ultra is

0:03:00.320 --> 0:03:06.400
<v Speaker 1>incredibly seamless. The same rascale architecture is going to be

0:03:06.520 --> 0:03:09.600
<v Speaker 1>used for Vera Rubin, and so the supply chains are

0:03:09.680 --> 0:03:13.760
<v Speaker 1>used to it this complexity that we enjoyed with Grace

0:03:13.840 --> 0:03:18.040
<v Speaker 1>Blackwell transition. We're now incredibly smooth running and so I

0:03:18.040 --> 0:03:20.600
<v Speaker 1>think Vera Rubin is going to be just really smooth

0:03:20.600 --> 0:03:21.960
<v Speaker 1>and we're gonna rt it really hard.

0:03:22.919 --> 0:03:26.120
<v Speaker 2>Jensen. I tried to go through what the CFO Collect

0:03:26.080 --> 0:03:29.720
<v Speaker 2>Crest said about China in the quarter gone It seemed

0:03:29.720 --> 0:03:33.120
<v Speaker 2>like there was not meaning for h twenty sales because

0:03:33.160 --> 0:03:35.880
<v Speaker 2>the demand wasn't there even if you were permitted to

0:03:35.920 --> 0:03:38.520
<v Speaker 2>sell each twenty. And then in the current period and

0:03:38.560 --> 0:03:41.720
<v Speaker 2>going forward in video seems committed to working with both

0:03:41.760 --> 0:03:47.120
<v Speaker 2>the United States and China to sell what Collect called

0:03:47.320 --> 0:03:50.960
<v Speaker 2>more competitive compute. Where do we stand with that and

0:03:51.320 --> 0:03:54.600
<v Speaker 2>could you just clarify what Collect was talking about in

0:03:54.640 --> 0:03:56.160
<v Speaker 2>the current state of play for China.

0:03:59.520 --> 0:04:03.200
<v Speaker 1>The most important thing she said is that we've said

0:04:03.280 --> 0:04:07.240
<v Speaker 1>for some time now our forecast for China is zero.

0:04:08.280 --> 0:04:12.360
<v Speaker 1>All of our forecast guidance that we showed zero, we

0:04:12.400 --> 0:04:15.080
<v Speaker 1>should start. That's the most important thing that she said.

0:04:15.360 --> 0:04:18.479
<v Speaker 1>She also said that effectively, China is a very important

0:04:18.480 --> 0:04:21.360
<v Speaker 1>market to us. It's very important to the United States,

0:04:21.600 --> 0:04:25.240
<v Speaker 1>it's very important to China. We would love the opportunity

0:04:25.600 --> 0:04:29.040
<v Speaker 1>to be able to re engage the Chinese market with

0:04:29.160 --> 0:04:32.159
<v Speaker 1>excellent products that we deliver and to be able to

0:04:32.160 --> 0:04:37.320
<v Speaker 1>compete globally. The Chinese market is very large this year.

0:04:37.600 --> 0:04:41.320
<v Speaker 1>My guess is probably about fifty billion dollars. It's great

0:04:41.360 --> 0:04:44.279
<v Speaker 1>for the American people that we're able to compete in

0:04:44.320 --> 0:04:47.240
<v Speaker 1>the Chinese market. It's great for the China market that

0:04:47.320 --> 0:04:51.080
<v Speaker 1>we're able to provide Nvidia's technology to them. It's great

0:04:51.080 --> 0:04:54.920
<v Speaker 1>for the rest of the world as Chinese software companies

0:04:54.960 --> 0:04:58.840
<v Speaker 1>and Chinese open source models leave China and are used

0:04:58.880 --> 0:05:01.720
<v Speaker 1>all over the world. And so I think it's fantastic

0:05:01.760 --> 0:05:04.200
<v Speaker 1>that we're able. It would be fantastic if we're able

0:05:04.240 --> 0:05:07.159
<v Speaker 1>to participate in the China market, but for now, we

0:05:07.160 --> 0:05:13.040
<v Speaker 1>should just assume our nvidious forecast for China market is zero.

0:05:13.360 --> 0:05:16.320
<v Speaker 1>We're going to continue to engage the US government, continue

0:05:16.360 --> 0:05:19.880
<v Speaker 1>to engage the China government to advise them and to

0:05:20.160 --> 0:05:22.080
<v Speaker 1>encourage them to allow us to go back and compete

0:05:22.120 --> 0:05:24.840
<v Speaker 1>in the open market. And so until then we should

0:05:24.839 --> 0:05:27.080
<v Speaker 1>assume zero Jensen.

0:05:27.160 --> 0:05:30.760
<v Speaker 2>During the call, the US Commerce Department issued a statement

0:05:30.920 --> 0:05:34.479
<v Speaker 2>saying that you are now permitted to export up to

0:05:34.560 --> 0:05:38.800
<v Speaker 2>thirty five thousand Blackwell chips each to both Saudiast humane

0:05:39.160 --> 0:05:41.920
<v Speaker 2>and to the UAE through G forty two. But there

0:05:41.960 --> 0:05:45.200
<v Speaker 2>are some requirements that the US has a view, in

0:05:45.240 --> 0:05:51.120
<v Speaker 2>particular around controls of preventing tech transfer to China through

0:05:51.120 --> 0:05:53.000
<v Speaker 2>the Middle East. What can you tell us about your

0:05:53.080 --> 0:05:55.480
<v Speaker 2>understanding of what the US government's asking of you.

0:05:55.520 --> 0:06:05.600
<v Speaker 1>There That that UH that that that element UH is

0:06:05.720 --> 0:06:08.880
<v Speaker 1>has been around for a long time is to prevent diversion.

0:06:09.640 --> 0:06:13.480
<v Speaker 1>Of course, over the years, people have speculated about diversion.

0:06:14.200 --> 0:06:19.720
<v Speaker 1>We've chased down every single concern and UH we've repeatedly

0:06:20.400 --> 0:06:24.320
<v Speaker 1>UH tested and sampled data centers around the world and

0:06:24.680 --> 0:06:28.520
<v Speaker 1>found no diversion. And so this is a this is

0:06:28.560 --> 0:06:31.320
<v Speaker 1>an area that that will continue to be rigorous on

0:06:31.920 --> 0:06:36.080
<v Speaker 1>and UH there's a lot of different ways to uh, comply.

0:06:36.560 --> 0:06:39.000
<v Speaker 1>And one of them, of course, is to have it

0:06:39.080 --> 0:06:43.000
<v Speaker 1>be run by American Cloud. Another way is just to

0:06:43.000 --> 0:06:46.040
<v Speaker 1>make sure that we have measures put in place, whether

0:06:46.120 --> 0:06:49.560
<v Speaker 1>technology or processes, to ensure that no diversion happens.

0:06:50.920 --> 0:06:53.320
<v Speaker 2>Jensen, the number one question I get for you is

0:06:53.400 --> 0:06:58.680
<v Speaker 2>always about energy. How severe is the energy shortage in

0:06:58.720 --> 0:07:01.520
<v Speaker 2>the context of AI expects and would you talk a

0:07:01.520 --> 0:07:05.600
<v Speaker 2>bit about power and whether power is a bigger constraint

0:07:06.120 --> 0:07:08.440
<v Speaker 2>for this buildout than the chips themselves.

0:07:12.240 --> 0:07:16.720
<v Speaker 1>When you're growing at the rate and scale of Nvidia,

0:07:17.600 --> 0:07:21.120
<v Speaker 1>Remember we're growing some sixty percent a year. Just quarter

0:07:21.200 --> 0:07:25.040
<v Speaker 1>to quarter growth of our company is ten billion dollars.

0:07:25.600 --> 0:07:28.280
<v Speaker 1>We grew an entire size of a company just in

0:07:28.320 --> 0:07:32.600
<v Speaker 1>one quarter. And so the scale and the rate at

0:07:32.640 --> 0:07:35.720
<v Speaker 1>which we're growing, everything's a challenge, which is the reason

0:07:35.720 --> 0:07:39.160
<v Speaker 1>why Nvidia has to be world class at our supply chain,

0:07:39.560 --> 0:07:43.720
<v Speaker 1>working with incredible providers and suppliers like TSMC and the

0:07:43.760 --> 0:07:47.040
<v Speaker 1>memory partners and all of our systems partners, but also

0:07:47.120 --> 0:07:53.360
<v Speaker 1>working downstream to work with energy providers, power generator companies,

0:07:54.200 --> 0:07:57.760
<v Speaker 1>all of the land power and shell providers, so that

0:07:57.800 --> 0:08:01.480
<v Speaker 1>we could make sure that as we launch into the marketplace,

0:08:01.520 --> 0:08:04.560
<v Speaker 1>as we deploy into the marketplace, land Power and Shell

0:08:04.800 --> 0:08:08.360
<v Speaker 1>will be ready for us. One of our great advantages

0:08:08.800 --> 0:08:11.840
<v Speaker 1>is that we have such a large network of go

0:08:11.880 --> 0:08:15.640
<v Speaker 1>to market. We're in every single cloud. Every single cloud

0:08:15.680 --> 0:08:18.200
<v Speaker 1>service provider is a customer of ours. We're in every

0:08:18.400 --> 0:08:22.440
<v Speaker 1>single GPU cloud, and so we have a large network

0:08:22.440 --> 0:08:25.040
<v Speaker 1>and not to mention OEMs, not to mention, all around

0:08:25.040 --> 0:08:29.280
<v Speaker 1>the world. Our customer base, our network of partners is

0:08:29.320 --> 0:08:32.520
<v Speaker 1>so large that we will find nooks and crannies of

0:08:32.559 --> 0:08:36.840
<v Speaker 1>power and large scale, medium scale, small scale in different

0:08:36.840 --> 0:08:39.400
<v Speaker 1>parts of the world. And so this is a huge

0:08:39.440 --> 0:08:42.160
<v Speaker 1>advantage of ours, and it stems from the fact ed

0:08:42.600 --> 0:08:48.600
<v Speaker 1>that Nvidia's architecture literally runs every model and today, yesterday

0:08:48.600 --> 0:08:51.800
<v Speaker 1>we announced a big news with Anthropic and so now

0:08:52.400 --> 0:09:01.720
<v Speaker 1>the premiere frontier models, open AI, Anthropic, XAI, Gemini, all

0:09:01.720 --> 0:09:07.280
<v Speaker 1>the open sources, biological models, physical AI models, everything in

0:09:07.320 --> 0:09:10.240
<v Speaker 1>the world runs on Nvidia. And as a result of that,

0:09:11.080 --> 0:09:16.680
<v Speaker 1>irrespective of which cloud provider you are, it is fantastic

0:09:16.720 --> 0:09:18.800
<v Speaker 1>that we can deploy in your cloud because the off

0:09:18.880 --> 0:09:19.880
<v Speaker 1>take will be incredible.

0:09:20.760 --> 0:09:23.640
<v Speaker 2>Jensen. We can see where the hyperscalers are getting the money,

0:09:23.679 --> 0:09:26.400
<v Speaker 2>where they have the money to deploy and build. But

0:09:26.640 --> 0:09:30.040
<v Speaker 2>you mentioned anthropic. With Anthropic or indeed open ai, they

0:09:30.040 --> 0:09:35.040
<v Speaker 2>have tens of billions of dollars of commitments around the world.

0:09:35.600 --> 0:09:37.640
<v Speaker 2>Very simple like, how do you know the open ai

0:09:37.760 --> 0:09:39.440
<v Speaker 2>is good for it that it will be able to

0:09:39.480 --> 0:09:42.920
<v Speaker 2>find the money.

0:09:43.840 --> 0:09:50.079
<v Speaker 1>Well, we're thoughtful along with open ai, thoughtful in aligning

0:09:50.160 --> 0:09:56.320
<v Speaker 1>on and taking into consideration visibility of demand and their

0:09:56.360 --> 0:10:00.000
<v Speaker 1>financing capabilities. All of that has to be in accordance,

0:10:00.559 --> 0:10:03.840
<v Speaker 1>has to be aligned, has to be coherent before we

0:10:03.880 --> 0:10:08.920
<v Speaker 1>start to build out. And so I think the ambitions large,

0:10:09.240 --> 0:10:13.199
<v Speaker 1>but the execution is disciplined, and that's really really important

0:10:13.200 --> 0:10:17.040
<v Speaker 1>to recognize. We're very disciplined with our investment. We're disciplined

0:10:17.040 --> 0:10:21.320
<v Speaker 1>with our buildout. These are very large scale investments, and

0:10:21.400 --> 0:10:24.920
<v Speaker 1>so the two teams are quite disciplined, very disciplined in

0:10:25.000 --> 0:10:28.559
<v Speaker 1>thinking through the investment levels. Now, it's also important to

0:10:28.600 --> 0:10:32.199
<v Speaker 1>take a step back and realize that open Ai, Anthropic,

0:10:32.760 --> 0:10:36.920
<v Speaker 1>these are the fastest growing company in the history of humanity.

0:10:37.440 --> 0:10:42.600
<v Speaker 1>They're off take their end market demand is absolutely real

0:10:42.880 --> 0:10:46.439
<v Speaker 1>and absolutely incredible, and you could see that they're really

0:10:46.480 --> 0:10:49.120
<v Speaker 1>struggling to keep up with the demand that they have.

0:10:49.679 --> 0:10:52.320
<v Speaker 1>The engineering teams. We work incredibly hard to make sure

0:10:52.360 --> 0:10:56.120
<v Speaker 1>that we bring them on more capacity, but also optimizing

0:10:56.160 --> 0:10:59.360
<v Speaker 1>their stack so that the usage of whatever capacity is

0:10:59.360 --> 0:11:03.320
<v Speaker 1>as sufficient as possible. And meanwhile, there's so many new

0:11:03.400 --> 0:11:05.679
<v Speaker 1>use cases that they want to put back put out

0:11:05.720 --> 0:11:09.040
<v Speaker 1>into the world, and it's currently limited by the capacity

0:11:09.120 --> 0:11:11.880
<v Speaker 1>they have, and so this is a really important time.

0:11:12.200 --> 0:11:15.439
<v Speaker 1>You're seeing an exponential growth in the amount of compute

0:11:15.520 --> 0:11:20.120
<v Speaker 1>demand necessary for AI, You're seeing an exponential growth of

0:11:20.400 --> 0:11:25.439
<v Speaker 1>adoption and use of AI, and the number of applications

0:11:25.480 --> 0:11:28.000
<v Speaker 1>that are going to be using these AI is also growing.

0:11:28.200 --> 0:11:30.400
<v Speaker 1>And so we've got to do our best to support

0:11:30.440 --> 0:11:33.560
<v Speaker 1>the scaling out of two of the most consequential companies

0:11:33.559 --> 0:11:37.320
<v Speaker 1>in history, and we're delied to be partnered with them.

0:11:37.840 --> 0:11:44.079
<v Speaker 2>Jensen investors have been worrying about depreciation. Software can actually

0:11:44.160 --> 0:11:46.880
<v Speaker 2>extend life, like there are a one hundreds out there

0:11:47.080 --> 0:11:51.720
<v Speaker 2>in the real world still at full utilization. Are people

0:11:51.800 --> 0:11:55.880
<v Speaker 2>underestimating how long your chips stay useful or are they

0:11:55.960 --> 0:12:00.080
<v Speaker 2>kind of misunderstanding in the context of depreciation, how oh

0:12:00.120 --> 0:12:03.400
<v Speaker 2>you're handling generation to generation updates of GPU.

0:12:06.920 --> 0:12:15.440
<v Speaker 1>Nvidia's architect Nvidia is unlike any other accelerator and The

0:12:15.559 --> 0:12:21.160
<v Speaker 1>reason for that is because of Kuda's diversity of capability

0:12:21.400 --> 0:12:25.320
<v Speaker 1>and versatility. Remember I said two things earlier. I said

0:12:25.320 --> 0:12:30.360
<v Speaker 1>the fact that Nvidia participates and could accelerate every phase

0:12:30.400 --> 0:12:34.240
<v Speaker 1>of AI, pre training, post training, and inference. Were the

0:12:34.280 --> 0:12:37.360
<v Speaker 1>only architecture in the world that does that fantastically. The

0:12:37.440 --> 0:12:41.720
<v Speaker 1>second thing we do, we run every single model, and

0:12:41.800 --> 0:12:45.840
<v Speaker 1>so most agentic systems, most clouds are running so many

0:12:45.880 --> 0:12:50.319
<v Speaker 1>different diverse type of models language models, vision models, biological models,

0:12:50.400 --> 0:12:54.000
<v Speaker 1>chemical models, for all the different fields of science. Nvidia

0:12:54.320 --> 0:12:59.640
<v Speaker 1>could be used across the entire lifespan of the technology.

0:13:00.120 --> 0:13:02.720
<v Speaker 1>And so if you look at products that we shipped,

0:13:02.920 --> 0:13:06.720
<v Speaker 1>Ampier A one hundred we shipped six years ago. But

0:13:06.800 --> 0:13:09.640
<v Speaker 1>because we're continuing, our installed bas is so large, our

0:13:09.679 --> 0:13:14.400
<v Speaker 1>diversity is so great, we could continuously update our software

0:13:14.920 --> 0:13:17.480
<v Speaker 1>bring value to our customers on the one hand, but

0:13:17.559 --> 0:13:22.240
<v Speaker 1>because our versatility is so great based on the capability

0:13:22.240 --> 0:13:24.640
<v Speaker 1>they need, they could use our GPUs for a very

0:13:24.760 --> 0:13:27.560
<v Speaker 1>very long time. Now, remember A one hundred and six

0:13:27.640 --> 0:13:31.160
<v Speaker 1>years old. However, it is still an order of a

0:13:31.200 --> 0:13:34.760
<v Speaker 1>magnitude faster than any CPU could put the bear. So

0:13:34.840 --> 0:13:37.240
<v Speaker 1>it is still the best computer. It is still the

0:13:37.280 --> 0:13:40.400
<v Speaker 1>best processor for much of the workload in the cloud,

0:13:41.000 --> 0:13:46.920
<v Speaker 1>and most people misunderstand that because unlike US, most accelerators

0:13:47.520 --> 0:13:51.239
<v Speaker 1>are kind of singular use, because they don't have diversity,

0:13:51.480 --> 0:13:54.320
<v Speaker 1>because they don't have versatility, because they're not great at

0:13:54.360 --> 0:13:58.200
<v Speaker 1>every phase of AI, once they're used for whatever they

0:13:58.240 --> 0:14:01.520
<v Speaker 1>were designed to do of value fast off a cliff.

0:14:01.800 --> 0:14:04.680
<v Speaker 1>That is not true with n video Jensen.

0:14:04.679 --> 0:14:08.040
<v Speaker 2>My final question is a point of clarification, if I may.

0:14:08.080 --> 0:14:11.319
<v Speaker 2>You were asked on the call about content and in

0:14:11.440 --> 0:14:15.040
<v Speaker 2>video's contribution to any given piece of AI infrastructure, and

0:14:15.080 --> 0:14:17.839
<v Speaker 2>what you said was Hopper around twenty to twenty five

0:14:18.320 --> 0:14:23.240
<v Speaker 2>Blackwell about thirty with those figures billions of dollars in

0:14:23.360 --> 0:14:26.760
<v Speaker 2>dollar terms on a one gig of what data center?

0:14:26.840 --> 0:14:29.320
<v Speaker 2>Or are you talking about percentages of total cost?

0:14:32.800 --> 0:14:38.160
<v Speaker 1>Oh, thank you, billions of dollars billions of dollars. Yeah,

0:14:38.200 --> 0:14:42.560
<v Speaker 1>And so for our ver Rubens system, a one gig

0:14:42.560 --> 0:14:45.400
<v Speaker 1>of what data center is probably something along the lines

0:14:45.400 --> 0:14:49.080
<v Speaker 1>of fifty to fifty five, and video's contribution is probably

0:14:49.080 --> 0:14:50.000
<v Speaker 1>about thirty five of.

0:14:50.000 --> 0:14:57.960
<v Speaker 2>That video CE one billion noted. I was asked to

0:14:58.000 --> 0:15:01.680
<v Speaker 2>ask you, the question has been asked answered astonishingly. Busy

0:15:01.800 --> 0:15:03.720
<v Speaker 2>day for you in the nation's Captain.

0:15:03.880 --> 0:15:07.680
<v Speaker 1>That's a great question. The difference between percentages and billion

0:15:07.800 --> 0:15:09.840
<v Speaker 1>is a big deal. Yeah, it's a brilliance.

0:15:10.560 --> 0:15:12.920
<v Speaker 2>The stakes are high, but we're always grateful for your time.

0:15:13.200 --> 0:15:15.000
<v Speaker 2>I know it's been a busy day. Thank you for

0:15:15.080 --> 0:15:18.360
<v Speaker 2>joining us on Bloomberg Television. That's Nvideo CEO Jens Wang