I don’t think it’s an understatement to say that Nvidia wants to be the company that drives the artificial intelligence universe that we might one day live in permanently.
The opening keynote speaker at the NVIDIA GTC 2025, which was mentioned jointly as Woodstock of AI, so CEO Jensen Huang elaborates on the virtues of an AI ecosystem driven by hardware – and software – from the second most valuable company in the world.
We were reintroduced to the concept of AI factories, which are essentially gigantic data centers, and where Nvidia’s world, the tokens generated corresponds to the revenue: The presentation mentioned a price, $ 1 trillion.
The cost of building data centers, like mega-yard, increases exponentially every generation when they become more complex, and require more power to house even more calculation.
The High Stargate project has a budget of $ 500 billion, and Apple has already committed to spending $ 500 billion over the next few years, a large part of which will be on data centers.
Depending on which analysts you are talking to, the global data center capital costs are expected to reach $ 1 trillion either in 2027 (PWC) or 2029 (dell’oro).
So it gives you a goal on Nvidia’s extremely ambitious goal for himself and for humanity.
A large piece of the puzzle that Nvidia introduced by the keynote speaker is its photonic switch systems that allow these AI factories to “scale to millions of GPUs”, apparently within the same physical perimeter.
Large, huge numbers
The largest number of them all in this presentation was when Pastor Lebaredian Nvidia’s VP, Omniverse and Simulation Technology, presented a slide on how physically AI transforms $ 50 trillion industries.
These span across two billion cameras, 10 million factories, two billion vehicles and Lebareian added ‘future billions of humanoid robots’.
Yes, it’s not just about robots, but humanoid robots, millions of them. Robots have generally existed for decades in many forms and forms.
Nvidia, however, imagines a world where billions of robots are built on what it calls, its “three computers”: before and after exercise (performed via DGX), simulation and synthetic data rating (performed via omniverse) and runtime (performed using AGX).
The world of the future may be home to “several types of humanoid robots” that will be able to post -train – while on the job, so to speak – with real or synthetic data so they can learn new skills or improve their knowledge base on the go, literally (or near) the speed of light.
And the first of them may well be based on its new GR00T N1 Humanoid Robot Foundation model developed in partnership with Google Deepmind and … Disney Research.
(Ed: You can’t see the bonding with GR00T from Guardians of the Galaxy and Wall-E, a potential prescient movie about the effect of robotics and AI on humanity).
Where does this leave us? With a clear timetable to a world where AI becomes an indispensable part of what it means to be human, perhaps the most important benefit of them all. Never, in the history of mankind, will so many will be dependent on so few.
Oh and the other acronym (AGI) were not mentioned during the presentation, not even.
