Nvidia (NVDA)'s first-quarter results beat analysts' expectations, with revenue up 262% to $26 billion. The company also announced a 10-for-1 stock split and a dividend increase.
In an exclusive interview with Yahoo Finance, Nvidia founder and CEO Jensen Huang spoke about the results and how demand for the company's products is “very strong.” He also mentioned how companies like Meta (META) and Tesla (TSLA) are pushing AI technology forward.
Jensen said Meta's Llama large-scale language model is “really, really important” given that it is “powering large-scale language modeling and generative AI efforts around the world.”
Regarding Tesla, Jensen explained that the company's latest fully self-driving technology is an “end-to-end generative model” that “learns by watching video and surround video and uses generative AI to learn how to drive.” I will.'' [to] Predict your path…how to understand and control your car. Therefore, this technology is truly revolutionary. ”
Watch the video to see why Huang says self-driving car AI systems are “most effectively trained by learning directly from video.”
Check out the full interview with Nvidia CEO Jensen Huang.
This post was written by Stephanie Mikulicz.
Learn more about Nvidia from Yahoo Finance below.
Nvidia shares rose 4% after the company reported better-than-expected profits, a stock split and a dividend hike.
Nvidia CEO Jensen Huang Named 'Person of the Year': Investor
Why this analyst says Nvidia isn't a stock to buy
How Nvidia's earnings will impact the chip market
Beyond the ticker: Nvidia
Video Transcript
Jess, I want to ask about cloud providers and the other industries that you mentioned, industries that are getting into the JA I game or getting NVIDIA chips.
You mentioned that in your comments in the actual release, and then I heard from your CFO, Collective Crescent, that 40% to mid-40% of your data center revenue is coming from these cloud providers.
These other industries are also beginning to develop.
What does that mean for NVIDIA?
Will cloud providers' share shrink? And will other industries take over the share of those cloud providers?
I expect both companies to grow in several different areas.
Of course, for consumer internet service providers, this last quarter was big news from Meta.
Well, the incredible scale of Mark's investment in Rama 2 was groundbreaking.
Rama 3 was even more amazing.
They create models that power large-scale language modeling and generative AI work around the world.
So the work that Meta is doing is really, really important.
Well, you saw Elon talking about the incredible infrastructure that he's building and one of the things that's really revolutionary about it is that with Tesla version 12, yeah, full self-driving is over. Ending generative models.
And it's going to watch the video, the surround video, and learn from that, and then, uh, drive it end-to-end and then use generative AI to generate, uh, the next path and its predictions, and, uh, learn how to steer, how to understand, and how to steer the car.
And this technology is truly revolutionary and the work they're doing is amazing.
So I gave two examples. A startup we work with called Recursion has built a supercomputer to generate molecules, understand proteins, and generate molecules for drug discovery.
The list goes on and on, and I mean, I could go on all afternoon and say that people from so many different disciplines are now realizing that we have software and AI models that can understand and learn just about any language, let alone English, and they can learn the language of images and videos and chemicals and proteins and even physics and generate just about anything.
It's basically like machine translation, and that capability is now being deployed at scale across a variety of industries, Jensen.
Just one more quick one.
Last question.
I'm glad you're talking about the car business. So what you're looking at is you mentioned that automotive is currently the largest vertical within the data center.
You talked about Tesla's business.
But what exactly is it?
Is that so? Do other automakers also offer self-driving cars?
What other features are automakers using within their data centers?
Please understand it a little better.
Well, Tesla is way ahead with self-driving cars.
But at some point, all cars will have to have self-driving capabilities.
It's safer, more convenient, and more enjoyable to drive, and it's very well known and well understood by now that learning directly from video is the most effective way to train these models.
Previously, we were training based on labeled images.
We say, this is a car, this is a car, this is a sign, this is a road, and we label it manually.
That's unbelievable.
And now we're putting video directly inside the car and letting the car figure it out for itself.
And the technology is very similar to that of large-scale language models, only it requires huge training facilities.
The reason is that we have video, the data rate of video, the amount of data of video is very high.
Well, the same approach used to learn the physical world from videos used for self-driving cars is essentially the same AI technology used to build large language models to understand the physical world.
Well, technology like Sora is really great.
Well, Google's other technologies have an amazing ability to generate meaningful videos conditioned by human prompts that need to be learned from the videos.
So the next generation of A has to be based on physical AI and it has to understand the physical world.
And the best way to teach these A's, how the physical world works, is through video, just watching loads of video.
This combination of multimodal training capabilities will therefore require significant computing demands in the coming years.
