SEO Dashboard
Subtitles
0:00Jensen won CEO of NVIDIA, fresh off his
keynote last night on stage. Great to see you.
Welcome to Las Vegas.
0:06Happy New Year.
Familiar territory. Congratulations on your new baby.
Thank you very much.
0:11A lot's changed since we last spoke,
actually. But it's the broad spectrum of what you
announced last night.
0:17New graphics cards, new chips, actually,
technically in the automotive space, products and services on the software
side.
0:24Which is the single most important for
Nvidia's future. They're all important.
You know, it's hard.
0:30It's single.
It's hard. It's hard.
It's hard to pick your favorites. You know, we announced three chips.
We announced a brand new A.I., a World
0:40Foundation model, and first of its kind.
And we announced our work in three areas
0:46in robotics.
Right. And they're all important.
And, you know, the the thing, of course,
0:52we announced a brand new Blackwell RTX
and has a new AI technology called
0:59neural negotiators.
And we're combining artificial
1:05intelligence and and classical compute.
Let's go to RTX, the new RTX. Yeah.
We've become so accustomed to your story
1:12being dominant in a market for high
performance CPU's server racks data centers.
This is going back to your roots.
1:19It's in the desktop context.
There's one right there later in the year laptop.
But for a target base that is developers
1:27and nerdy hardcore gamers.
What's the future of that business for you?
I computer graphics is going to be here
1:33forever.
Forever. And
what we've done is we've fused artificial intelligence and computer
graphics together, and the images that
1:42we're generating today wouldn't be
possible if not for the fact that we're using computer graphics to create
beautiful pixels and then use artificial
1:50intelligence to amplify that capability.
For example, you know, out of four
1:56frames I was talking about yesterday, 33
million pixels to sell in 4K. Now, out of that 33 million pixels, 2
million pixels were computed.
2:05The other 31 million pixels were
generated by AI. In other words, the AI predicts what it
thinks.
2:10The pixels.
That's right. Yeah.
The ultimate generative A.I.. But what was interesting for me is the
focus again was away from the graphics
2:18cards of way from Blackwell to underpin
nets and Cosmos. We probably don't have time to explain
in full detail Cosmos, but you call it a
2:25world Foundation model.
Cosmos is for the physical world. What Chad shipped is four words and
text.
2:34That's the easiest way to think about
that. Okay, so this model text input bots can
generate synthetic data in multiple
2:42mediums.
It understands the physical world. So, for example, if I ask you the
question, if I ask you to generate
2:50multiple features of a car driving down
the road, it would understand the
2:56dynamics of the world.
It would understand the object permanence and would understand
understand geometry in space, and it
3:03would create a driving scenario for the
car that is plausible.
3:09And so and you open sourced it.
So I don't really think about it as a product or a go to market.
It's more about what Cosmos enables.
3:18Is that how we should think about it?
Yeah, Well, the autonomous vehicle industry and the robotics industry is
really important to us and we offer
3:25three computers for them.
We offer it, of course, the training computer through through TJX, the
robotics computer that's inside the car
3:33or inside of Roman.
And now we have this new computer called Omniverse with Cosmos that is the
digital twin or the playground where
3:42these robots can learn how to be robots.
And so if we could accelerate the development of artificial intelligence
for Avs and for robotics, it brings in a
3:51lot of business for us.
Of course, there's an academic point of tension here.
If I may, Elon Musk is a customer of yours and Tesla.
The theory or practice is based on real
4:02world data gathered through vision.
Yeah. Does the synthetic data underpinning of
Cosmos kind of contradict that?
4:09It doesn't replace.
It augments. And so you're going to you should
collect as much world data as you can.
4:14Of course, collecting world data is very
expensive and Elon has a great advantage because, number one, he has his own
factory for his cars, is fantastic, has
4:24a lot of video gear in it.
His AVI algorithms is incredible.
4:29It's the best in the world and and he
has a very large fleet of cars on the road that allows him to collect a lot of
data.
4:36And so so I think he has just a
phenomenal position and he's been working on this for a long time.
And so he's he's going to be in a great
4:43position to to take advantage of it.
Well, may I ask you, at this juncture, he's clearly influential in this
upcoming administration, but he also
4:51positions Tesla as a leading air and
robotics company. Yeah, very.
How does that vote for him? Video Elon Musk's influence on President
elect Trump and.
4:59And also the incoming administration's
kind of attitude towards I,
5:05I don't know, the the the the attitude
towards it. I, I know Elon's attitude towards the
and and he's very optimistic about the
5:13future.
And obviously he's working on some of the most important areas.
He's working on foundation cognitive
5:22intelligence i i.
His. Tesla is working on Thomas vehicles and
optimizes for him nor robotics.
5:32These three areas of A.I.
are the three most important areas of A.I..
And so I think I think he's working on
5:37exactly the right things.
You kind of positioned and video is positioned in the supply chain for
physical A.I.
5:44robotics, autonomous driving.
Explain that a bit more, the role you see in video play.
Well, we're a core technology company,
5:51and so we build the foundational
computing platforms. We're also full stack.
And so we developed the necessary
5:57algorithms and necessary A.I.
technologies, and then we put it and we put it out to the industry for them to
adopt it and turn it into end market
6:05solutions where we're computing platform
companies. So you're on stage and you're surrounded
by, I think, a dozen humanoid robots,
6:13which is nice for you.
When will I be here at CES in Las Vegas and actually have there some robots
right now in real terms?
6:21You must have a timeline that you see
real world commercial deployment of the
6:26technology you outlined last night.
It depends on on use case. I would say first use the first use case
will probably be in manufacturing.
6:34You know, there's
estimates tens if not 100 million jobs
6:40that are workers that are the world is
short of workers and
6:46aging population declining declining
birth rates. And and so I think the the world needs a
lot more workers.
6:55Robotics is one of the best ways for us
to supplement all of that and help
7:00companies recover the lost revenues on
the one hand and drive productivity, which reduces inflation for the world.
On the other hand.
7:08And so I think robotics is going to be
very important to that in different different areas.
So you could have probably deploy to
7:14manufacturing first because they
obviously need it most, which you see is a ginormous potential market
addressable.
7:21It's a $50 trillion industry that wants
to grow and things to grow. Sorry to interrupt you.
Jensen.
7:26It's interesting that Wall Street is
struggling with this morning, among many things, is
they understand investment there that
7:34drains the foundation models.
In terms of Nvidia's business model, what you outlined last night, Cosmos and
then later on, on the inference side,
7:45how physical AI helps grow your business
from that drives growth. Just simple as that.
Yeah, as simple as that.
7:51I think if you if you just look at
simply like we have three computers and two of the computers and omniverse
drives an enormous amount of data that
8:00is necessary to train the A.I.
models. Right?
And so Omniverse creates the data that
8:05we then use to train A.I.
models. The training is what drives direct
sales. And the more robots that are that are
available, the more data we can create,
8:15the more A.I.
models we have to. We have to go train.
That's that cycle is ultimately what
8:21we're striving for.
All of that drives consumption from data center growth.
There is a surge in AI spending, data
8:28center growth.
Mm hmm. Some of our audience are a bit concerned
about how sustainable that is.
8:33Short, medium and long term.
Well, at the at the limit, artificial
8:39intelligence is the single most
important technology force of our time. And and it's really about we're at the
beginning of that.
8:46And in the future, every single data
center will be driven by AI and and the type of
computing that we build today.
8:53And and so if you look at the world
today, we're about a year and a half into the remodeling, if you will, the
modernization and the reinvention of
9:03computing.
And so I think that over the next several years, you're going to see the
transition from the old way of doing
9:09computing, general purpose computing.
This is a new way of doing computing, artificial intelligence and accelerated
computing.
9:15And so we have a lot of growth to go to.
Let's go back to your and Nvidia's roots, and therein lies the complication
behind the story Accelerated computing.
9:24You spent four years redefining the
computer for me, but then we get RTX
9:30Blackwell single.
BLACKWELL In that form factor, the target audience is hardcore gamers, but
also developers.
9:38And I wondered if you could give me any
early examples or evidence of how you see the gaming industry adopting your
technology.
9:46Well, A.I.
is going to re reinvigorate the video
9:51game industry.
On the one hand, for developers, it's going to reduce the cost of creating the
content.
9:57On the other hand, all of the characters
that are in the games are going to be smart characters in the future.
So every time you interact with them,
10:04they're going to be interacting with you
in a much more intelligent way. And so the games are going to be more
interesting, the characters are going to
10:10be more interesting, the content
development costs is going to decline, and that's going to be really great for
the industry.
10:17And so I think the future is really
bright for for video games and in these virtual worlds.
And artificial intelligence is going to
10:24really.
Project. Digits.
May I pick it up? Yeah, yeah, yeah.
$3,000.
10:31Yeah.
Digits? Yeah.
A supercomputer. Yeah, exactly.
How do you imagine you're just sitting
10:37there just like that?
You're working on your PC? Well,
why would I need one of these?
10:43Probably not me.
But how big is the addressable market for for this?
What is the addressable market?
10:49There's 30 million software developers
there. Probably something along the lines of 10
million designers around the world.
10:57Probably another 20 million creative
artists. Hard to say exactly how many students
I'm going to guess.
11:04Probably a couple of hundred million
students around the world. Everybody is going to have to at it.
Everybody's going to have to to what
11:12they can afford computers.
And so here's if they can afford computers and they would like to have a
companion that helps them, do I?
11:21This is the way to do it.
Can I just clarify something? I think you said on stage Mac OS Linux
and Windows, But no, no education, said
11:30Jocelyn.
No, no. Whatever computer you use, you're
literally enjoying how it's going to be
11:36used.
It's sitting right there and you're just connect to it wirelessly like it's your
personal cloud.
11:42I promised the audience I'd clarify
that. Yeah, yeah, yeah, yeah.
We're running short on time. President elect Trump has been speaking
during the course of my conversation.
11:51How imperative is it that you go
tomorrow? I'll go and meet with him.
If Nvidia is America's leading air
11:57company.
And will you? I'd be delighted to go see him.
Have you been invited? Not yet, but I would be delighted to go
see him and
12:06and congratulate him and and do
everything we can to help this administration succeed.
A lot of what you outlined on stage last
12:13night in the realm of fiscal I you know,
I saw xpeng, for example, in the autonomous driving context that's
happening in China, Like they are doing
12:21a lot on robotics.
So I'm going to ask you about tariffs. You know, it's likely this coming
administration will be as restrictive on
12:28technology exports and tariffs will be a
function. How have you prepared for that?
Jensen?
12:35Well, whatever whatever the
administration ultimately decides, we'll give them as much insight as we can from
our perspective.
12:43And I'm sure that the administration
will make the right moves that's in the best interest of our country.
2025.
12:50I kind of started the conversation by
saying a lot has changed since we last spoke in the summer.
You said that the age of general
12:57robotics is just around the corner.
Is 2025 the year of general robotics, or
13:02is that a little premature to your mind?
The development is going going gangbusters, as you can see, all the
different robots that are going to be
13:10around here and the enabling technology
necessary for general robotics is coming
13:15together.
All the pieces are coming together. The industry still has a lot of
engineering to do.
13:21If you if you look long term, you know,
pick your horizon in ten or 20 years,
13:26the number of robots that's going to be
on Earth is going to be measured in probably tens, if not hundreds and
potentially billions of robots.
13:35And so those those days are clearly
coming. Is it going to happen in the next couple
of years or the next five years?
13:42Hard to say, But the development of
robotics is going to be all over the world now.
And we're seeing startup companies,
13:48large companies, industrial companies,
consumer electronics companies all getting involved in the future of
robotics.
13:54And our offering to the industry is a
three computer system. And so whether they're developing the
robot training, the robot, we have
14:01systems for them and GE X clouds for
them. If they're simulating the robots, we
have omniverse for them.
14:07And if they want to deploy that when
they're ready to deploy the robots, we have, you know, little or little,
little, little computers that basically
14:14is the computer brain of the robot that
they can put inside a robot. And so we'll work with the industry
across the board from the development of
14:22the world blind to the deployment of the
robot. And we have computer systems for them,
algorithms for them, for them, and we'll
14:30partner with the industry to make this
future happen very, very quick. Which line of business grows fastest
this year?
14:35Datacenter, gaming or other, they're all
going to grow fast. I think gaming is continuing to grow.
Our our autonomous vehicle business is
14:45already on its way to be a $5 billion
business just this year. And so right and right run rate.
And so the autonomous vehicle business
14:54is is just starting to get off the
ground. And that tells you something about how
we address it. And the reason for that is we get the
benefit from the beginning of the
15:02development of the AIS all the way to
the deployment of the cars, because a car company needs to factories, a car
factory that builds the cars and an AI
15:12factory that builds the AIS for the
cars. And both of these the both of these
areas, we get to participate.
15:18And so I think it's going to be a very
large business. Jensen I've made you late for your next
appointment. Yeah, it's great to see you.
Really grateful for your time.
15:24It's good to see you.
While Jensen won the immediacy.