- 10 hours ago
Category
🗞
NewsTranscript
00:00I want to try and understand something you mentioned very briefly, which is the strategic
00:04importance of not just the equity stake, but the partnership on technology and not just for
00:09NVIDIA, but for America. Why are you moving into this domain? One, we announced a brand new platform
00:16for a new market we've never been in. And so that's number one. Number two, if we have,
00:21if we take advantage of this transition from general purpose computing to accelerated computing,
00:25from old types of software to AI, we can take advantage of this transition of AI and 6G
00:33for America to win back telecommunications again. It's been a long time since the American
00:38telecommunication network has been built on American technology. You used the word national
00:42security, I believe. For national security reasons, for economic reasons, our industry should be built
00:46on American technology. And for the very first time, we can do that. We have a brand new product line
00:50that takes advantage of computing, accelerated computing and AI. And we have a great partner
00:55to help us do it with Nokia. The market reaction kind of speaks for itself. There was a time when
01:02Nokia was already looking to America. You've been in the role since April. But may I start by asking
01:07you what role the administration played in bringing you both together if this is indeed a strategic
01:12priority? Well, I don't think there was anything specific that the administration did other than
01:17they've created the environment to support innovation. I mean, Jensen's talked about
01:21manufacturing, but it's also about R&D and innovation and technology leadership. And if you
01:27think about where this world is going and all those incredible devices that are going to be built on
01:31NVIDIA platforms, robotics, autonomous vehicles, augmented reality, virtual reality, we need a different
01:37kind of network in the future. And that's what we realized. And that's why we wanted to forge this
01:42partnership so that we're building a network leveraging AI for AI services. And that's really
01:49the big change here. And the technology is flowing both ways. I mean, what is it that you will be able
01:53to do with Nokia, Jensen, that you weren't able to do on your own? Well, first of all, Nokia is in all
02:00of the world's base stations. And this air scale platform of Nokia is in millions of base stations
02:05around the world. If our computers are not inside that base station, it doesn't help. It's no different than
02:09NVIDIA's tech computing platforms, not inside a car. It doesn't help. And so the first thing is we
02:14need a partner to get this into the world's space stations. We also, remember, we're bringing AI to
02:20radio networks, AI to RAM, so that we can make wireless tech communications a lot more efficient.
02:26Second, we're bringing AI for the radio, meaning on top of these radio networks and these telecommunication
02:32network, we're going to be able to provide AI services, which is going to be able to make it easier
02:37for us to do robotics and autonomous vehicles and industrial automation and reach all of the
02:42different places around the world. That's kind of hard to do with Wi-Fi. Cellular reads everything.
02:48And so we now have this fabric that we can deploy AI computing services on top of. It's going to be
02:55completely revolutionary. In your world and in your market, Huawei is an influential player, not just
03:02in Europe. I think also about the Middle East and Africa. How important is the relationship with
03:08NVIDIA to be able to counter, you know, what Huawei does in those markets?
03:13Yeah, look, I think for us, one of the principles I've had with my team since coming in has been
03:17do what we can do that no one else can do, and that is air scale, what we do with any RAN software,
03:23the software that runs on it that we're going to be putting on the NVIDIA ARC platform.
03:28That do what we can do well and partner with the best everywhere else. And this is a this is a clear
03:34opportunity to partner with the best. And by the way, that's where America and the Western world
03:40succeeds, because we don't necessarily have one company doing everything like Huawei. We succeed
03:45because we partner aggressively. And you if you watch GTC today, it was hard to not find it obvious that
03:51you're partnering with everybody. But that's that's the advantage is this partnership. So we do what we do
03:56best. We let NVIDIA do what they do best. And ultimately, our customers win. And ultimately,
04:02for America, I think America will win because there'll be more innovation faster.
04:05Nobody knows telecommunications better than Nokia. They've been at it for a long time. They're all
04:11over the world. And between our two partnerships, we now have wireless telecommunications. We have AI
04:16and we have accelerated computing all in this partnership. And so all of that capability is
04:22going to be brought to bear for the next generation of 6G. And I think that's this is a this the
04:28capabilities we're bringing together in this partnership is quite unique and
04:30informative and form formidable. Why was it important, Jensen, that NVIDIA ends up with
04:36almost three percent of Nokia? What was the rationale behind that? The work that we're doing is so
04:41intensive. And quite frankly, I was so optimistic about it and so excited about it and that we wanted to
04:46invest in Nokia so that we could be part, you know, of this incredible success that we're creating
04:51together. And I got to tell you that that move was pretty genius, wouldn't you say?
04:55That's not for me to comment on. What I'd observe is, you know, a 25 percent jump in the ADRs.
05:01What does that signal to you? You know, NVIDIA, with respect, Jensen, I've known to be about speed.
05:07Is it that they're giving you a technology that you wouldn't otherwise have or have you now a way of just
05:11moving faster? Yeah, it's both. I mean, there's obviously, you know, there's no way we're going to
05:17have NVIDIA's technology and accelerated computing. I mean, the CUDA software stack, all of the
05:21capabilities, even the fundamentals of what's in Arial, the software stack that NVIDIA has built is
05:27an accelerator for us. And then it's about speed because now we can innovate much faster. And for us,
05:33it's that shift into adding value in software and thinking more like a software company or really not
05:38like a software company, like an application company that's going to use AI and AI models
05:42to continue to innovate and move faster. And I think the market recognizes that this is the kind
05:47of innovation that this industry has needed. Look, this industry has a history. 3G, we were building
05:53better voice networks when the internet was getting started. You know, 4G, 5G, we were building data
05:57networks behind where the market was. And today, 5G, when you look at it globally, is still only partially
06:03deployed. So as you look ahead, in order to innovate, you've got to move much faster. And we need to
06:08enable our customers to have different and better technology. And with this partnership, all of a
06:13sense, speaking of speed, we're in Blackwell today, it's going to go to Rubin, it's going to go Feynman,
06:18and it's going to go on a clip that basically is at the speed of light. I've got to ask you to
06:22clarify something, Jensen. You showed a slide behind you, $500 billion over what I think illustrated
06:28six financial quarters, but 20 million Blackwell. Five financial quarters. Five financial quarters.
06:35Does that figure include networking? We know it doesn't include China, because it seemed to be
06:40technically above consensus, right? But you seem to be suggesting that there is near term, actually,
06:45an acceleration beyond what you were expecting. All of that was only data center. That's the amazing
06:51thing. All of that is only data center. Today, we announced several new platforms,
06:54the NVIDIA drive Hyperion platform, our Arc platform, our quantum platform. None of that is
07:01included in that half a trillion dollars. No networking? No networking whatsoever. Well,
07:06that's data center. So networking, NVIDIA networking for data centers, but not wireless networking,
07:10not wireless telecommunications. When do we see a real-world manifestation of the work that you're
07:15both going to do together? Yep. There's work going already. You can see a demo, actually, at the booth
07:19here in GTC. You'll also see customer trials early next year, and then we think in full commercial
07:26production in 27. And, you know, the great thing about the partnership with Jensen is I know he'll
07:31be pushing me to go faster, and I'll be pushing his team to go faster, and together that means that we
07:35can get to market as soon as practical for our customers. And Arc is backwards compatible with
07:41air scale. Right. And all of those millions of base stations could be upgraded to 6G and AI.
07:46There's a lot of skepticism on 6G, though, right? Like, we're agreeing standards. Just a simple yes
07:51or no question, really. Is the 6G timeline accelerated because of this partnership?
07:56I think it is, because what you can actually do now is, you know, exactly to what Jensen said,
08:00you can deploy with Arc, and then if something happens or the standard evolves or, you know,
08:05or there's changes, it'll just be a software model update that we deploy in applications.
08:11It's a completely different model than what we're used to back in the days of deploying these
08:15complex integrated legacy appliances. It's general purpose accelerated computing,
08:19and that's why we think it's so compelling. If the standards change, you got to go fix a chip.
08:23If the standards change, you're just, and I just upgrade the software, the CUDA software. That's the
08:27amazing thing. Jensen, you are traveling onwards now to Asia. I don't know what your schedule is for
08:34the jet, and I don't know if you and the president will be at the same place at the same time. He seemed
08:37to think you would be. But the main point is, is that you're here in Washington, D.C. We're a few blocks
08:43from the White House. You called it the AI Super Bowl. Others called it that too. But everyone is
08:47inferring from this that this was about all of the things we talked about all year, that you want to
08:54be closely aligned with this administration. And while in Asia, that there is something that the
08:59president needs to work on in the context of China trade negotiations. Whether you get to meet him or
09:04not, what is it that you need the president to do for NVIDIA to accelerate on its plans?
09:09We need the president to do what's right for America. And I'm certain he's going to go do
09:13that. The reason why we're here is because technology is now so important in politics and
09:18geopolitics. It is the single most important industry in America. It is our national treasure.
09:24And it's important for us to make sure that Washington, D.C. and the policy makers and all of
09:31the people who serve our country to make it great understands the technology as it stands today and where
09:37we're going to go and and the companies that matter to us so that we can make America great.
09:41And so that that was the primary reason as we as well, first of all, whenever there's an opportunity
09:49to see the president, I always love that. He's he's incredibly engaging. He's always learning and he
09:54wants to help America win. He wants America to win. This is my final question. I was trying to
09:58interpret what you were saying. It was a detailed presentation. But you talked about how AI is now
10:04worth paying for. You talked about the end or not of Moore's law, depending on your point of view,
10:10but the need to scale to meet demand. Was that you saying we are not in an AI bubble?
10:17I don't believe we're in an AI bubble. And the reason for that is we're going through a natural
10:22transition from an old computing model based on general purpose computing to accelerated computing.
10:28We also know that AI has now become good enough because of reasoning capability,
10:33research capabilities, its ability to think it's now generating tokens and now generating intelligence
10:39that's worth paying for to the point where I'm paying lots of it. Cursor is all over NVIDIA.
10:45Every single one of our engineers use it. We pay lots and lots of money for it,
10:48and we're delighted to do it. And Cursor is just one of the applications. You know,
10:52what is it? 11 Tech. I forget. But anyways, all of these different AI models that we're using,
11:03we're using plenty of services and happily doing it.
Recommended
1:59
|
Up next
12:01
14:01
2:20
2:29
19:22
1:27
2:13
0:50
8:44
2:51
1:01
Be the first to comment