I do not think it is euphemism to say that Nvidia wants to be the company that drives the universe of artificial intelligence in which one day we could live permanently.
The key opening in Nvidia GTC 2025, colloquially known as Woodstock of AI, saw the CEO Jensen Huang exolt the virtues of an ecosystem of the Hardware and Software, of the second most valuable company in the world.
We reintroduce to the concept of AI factories, which are essentially gigantic data centers and where, the world of Nvidia, the tokens generated are equivalent to income: the presentation mentioned a price, $ 1 billion.
The cost of building data centers, such as Mega-Farms, increases exponentially in each generation as they become more complex and require more power to house even more calculators.
The Lofty Stargate Project has a budget of $ 500 billion and Apple has already promised to spend $ 500 billion in the coming years, a large part of which will be in data centers.
Depending on the analysts with whom he speaks, the capital expenditure of the global data center is expected to reach $ 1 billion in 2027 (PWC) or 2029 (Dell'oro).
That gives you a measure of the extremely ambitious objectives of Nvidia for itself and for humanity.
A large piece of that puzzle that Nvidia introduced into the key note are its photonic switching systems that will allow those AI factories to “climb millions of GPU”, apparently within the same physical perimeter.
Big, large numbers
The greatest number of all in this presentation was when the vice president, omniverse and simulation technology of Rev Lebradian Nvidia presented a slide about how physical AI is transforming $ 50 billion of industries.
These cover two billion cameras, 10 million factories, two billion vehicles and Lebareian added 'future billion humanoid robots'.
Yes, it's not just about robots, but humanoid robots, millions of them. Robots, in general, have existed for decades in many forms and forms.
However, Nvidia foresees a world on which billions of robots are based on what he calls, his “three computers”: PRE and post-training (made through DGX), simulation and generation of synthetic data (made through omniverse) and execution time (performed with AGX).
The world of the future can be the home of “multiple types of humanoid robots” that can post -Trenar, while at work, so to speak, with real or synthetic data, which allows them to learn new skills or improve their knowledge base on the march, literally (or close) of the speed of light.
And the first of them can be based on its new GR00T N1 Humanoid Robot Foundation model, developed in association with Google Deepmind and … Disney Research.
(Ed: one cannot see the link with Gr00t from Guardians of the Galaxy and Wall-E, a potentially prophetic film about the impact of robotics and AI on humanity).
Where does this leave us? With a clear road map to a world where AI becomes an indispensable part of what it means to be human, perhaps the most important utility of all. Never, in the history of mankind, so many will depend on so few.
Ah, and the other acronym (AGI) was not mentioned during the presentation, or once.