All eyes were on Nvidia (NVDA -0.32%) CEO Jensen Huang on Wednesday as he delivered his keynote speech at the 2025 CES in Las Vegas. The address was filled with several exciting announcements and it included some valuable insights from one of the foremost tech leaders, including some general discussion on artificial intelligence (AI) as well as how AI might affect Nvidia over the next few years.
The CES event helped show the investing world that there is still much to come from Nvidia and the AI industry as a whole. With all of the incredible growth already seen over the last few years, it wouldn’t be surprising to hear speculation about how much further the technology can go. If Huang’s comments on Wednesday are to be believed, the answer is a whole lot further. That bodes well for Nvidia shareholders.
Nvidia announces its newest consumer lineup
Nvidia’s focus of late has been dominated by data centers and AI models, but the heart of the company has traditionally been consumer graphics. The company was founded to help make graphic-intensive video games possible on personal computers and in a nod to this history (and the fact that CES got its start as the consumer electronics show) the first big announcement came in the unveiling of Nvidia’s newest line of RTX GPUs and their price points.
The newest line, based on Nvidia’s new Blackwell architecture (which also powers its AI chips), should at least double the performance standards of its current lineup, depending on the model chosen. That’s a significant upgrade considering its current generation is already among the best on the market.
Huang spoke about the integration of AI into computer graphics and how these performance increases were possible only because of AI-specific technology. Time will tell how they stack up in real-world testing once they are released, but Huang and Nvidia have a history of delivering on their claims.
Agentic AI is here
Huang took the opportunity to address what he called the next wave of AI, agentic AI. Where current generative AI excels at creating — text, music, even video — it cannot do all that much. That is, it can’t complete multistep, multivariable tasks on behalf of the user. That is exactly what agentic AI does.
As has been Nvidia’s primary model in the past, the company won’t be designing AI agents to be used directly by consumers or clients, but rather the building blocks that allow them to be developed and implemented using the Nvidia ecosystem. This allows Nvidia to focus on improving the underlying technology rather than worry about optimizing its use for specific applications. The model has been extremely successful in the past and it would seem it will be here, too.
Physical AI is coming
While agentic AI can be very effective, it is confined to the digital realm. Its application in the real world is the ultimate goal for Nvidia, according to Huang. This marriage of robotics, AI, and advanced sensory equipment could have massive economic implications. Think of the impact of truly self-diving vehicles or robotic assistants that can run errands for you.
This ultimate goal requires significant advancements in technology. A whole lot of training needs to take place. Traditionally, this training mostly comes in the real world. How else would you get the data? Now, Nvidia says it has the solution: Nvidia Cosmos.
This “world foundation model” uses AI to synthesize real-world environmental data that can be used to train AI intended for use in the real world. This could dramatically reduce the costs and time involved in training and testing robotics and greatly increase their success once placed in physical environments.
There’s a lot of flash at CES
Huang’s presentation certainly had a wow factor to it. It was designed to drum up excitement for the year ahead. Some of the specific claims and predictions may not come to fruition in 2025, but there was plenty of substance in the presentation, too.
The important parts of the Blackwell rollout, the maturation of agentic AI, and other comments all point to a strong year ahead for the company. They should all help the stock continue to outperform the market and add to the case for buying Nvidia.
Johnny Rice has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Nvidia. The Motley Fool has a disclosure policy.