Nvidia is a “technology company,” not a “consumer” or “enterprise” company, as CEO Jensen Huang emphasizes. What exactly does he mean by that? Nvidia doesn't want consumers paying hundreds or thousands of dollars for expensive new machines. RTX 50 Series GPU– They don't want companies to Buying more AI training chips? Nvidia is the kind of company that has a lot of know-how. To hear Huang tell it, if those pie crusts were the company's chips, AI would be the complement.
“Our technological influence will impact the future of consumer platforms,” Huang, wearing a typical black jacket and warm AI chest, said in a Q&A with reporters a day after his explosive CES opening. But how does a company like Nvidia fund it all? Those great AI experiments– H100 AI training chip Making Nvidia a technology powerhouse in the past two years as well A little stumbling Along the way, however, Amazon and other companies It is trying to create alternatives to reduce Nvidia's monopoly. What if competition reduces the fun?
“We will be responsive to customers wherever they are,” Huang said. Part of that is helping companies build “AI agents,” aka multiple AI models that can perform complex tasks. This includes several AI toolkits built to help businesses, even though H100 made Nvidia big and RTX keeps gamers coming back. But I want a new one. $3,000 AI processing center “Project Digits” To open up a “new universe” to those who can use it. Who will use it? Nvidia says this is a tool for researchers. Scientists and students, or at least those who happened upon the $1.50 cup of instant ramen they were eating for dinner for the fifth night in a row.
Nvidia Make sure you know Regarding the RTX 5090's 3,352 TOPS AI performance, Huang's company then dropped details about several software initiatives. both related to games and non-games None of his announcements was more confusing than his “foundation of the world” AI models. These models should be able to be trained in real-world environments. This could be used to help autonomous vehicles or robots navigate their surroundings. It's a lot of futuristic technology, and Huang admits he failed to explain it more clearly to the crowd that mostly came to see the cool new GPUs.
“(The Earth Foundation Model) understands things like friction, inertia, grab, the presence of objects. and components geometric and spatial understanding,” he said. “You know what kids know. They understand the physical world in a way that language models don't.”
Huang kicked off CES 2025 on January 1 with a packed keynote on the Michelob Ultra stage in the Mandalay Bay Casino in Las Vegas. Surely there were a lot of gamers who wanted to see the RTX 50 series cards. The latest in real form But there's more to see how the company fares as Nvidia moves forward. RTX and Project Digits drew cheers from the crowd. Half the time is spent talking about his model of the world's foundations. It makes the audience seem less enthusiastic.
It points out how awkward AI messaging can be. Especially for a company that is so popular among attentive PC gamers. There's been so much talk about AI that it's easy to forget that Nvidia was in the game for years before ChatGPT appeared. Nvidia's in-game AI upscaling technology, or DLSS, has been around for nearly six years. and is constantly being improved And now it's one of the best AI scaling tools in the game. Although limited by the exclusivity of Nvidia's cards, it was a good thing before generative AI was available. Nvidia now promises that the Transformer model will further improve upscaling and ray reconstruction.
Moreover, A touted multiframe generation could offer a fourfold performance boost for 50 series GPUs, at least if games support it. That's a boon for those who can afford the new RTX 50 series. The RTX 5090 tops out at $2,000. The gamers who will get the most benefit from next-generation frames are those who can only afford low-end GPUs. Huang declined to give any advice on the RTX 5050 or 5060, joking: “We have already announced four cards. And do you want more?”
The global foundation model is just a prototype. The same goes for Nvidia's new AI software on public display. The real question is When will it be ready for prime time? And who will use it in the end? Nvidia has shown NPC AI strange child– In-game chatbot– AI nurseand sound generator Last year, this year it wants to blossom with a global foundation model plus a host of AI “Microservices” It includes a weird animated talking head that's supposed to act as your PC's always-on assistant. Perhaps some of this will stick, in case Nvidia hopes AI will replace nurses or audio engineers. We hope that doesn't happen.
Huang considers Nvidia to be a small company with 32,000 employees worldwide. Yes, that's less than half the employees that Meta has, but you can't assume that would be negligible in terms of its influence on the market for AI training chips, given its official position. The market dominates the tech industry. The more people use AI, the more people will have to buy dedicated AI GPUs and other AI software. If everyone buys it themselves Home AI processing chipWithout the need to rely on external data centers and external chatbots, Nvidia is like any other tech company. You just need to find a use for AI. In addition to replacing all our work–