Jensen Huang says NVIDIA AI chips advance much faster than Moore’s Law

Jensen Huang, the jacketed CEO of NVIDIA, declared in a recent interview that his company’s AI chips are advancing “much faster than Moore’s Law”. This is an idea that the executive mentioned during his keynote for CES 2025, when the RTX 50 were revealed, but which he developed further by talking to TechCrunch.

“Our systems are progressing much faster than Moore’s Law,” said Huang. “We can make the architecture, the chip, the system, the libraries and the algorithms all at the same time. If you do that, then you can move faster than Moore’s Law, because you can innovate at every step.”

Moore’s Law is based on a prediction by Gordon Moore from 1965. The co-founder of Intel declared at the time that the number of transistors in computer components would more or less double every year, almost doubling their performance and reducing costs.

The “law” has proven true for years, although it has slowed down more recently. This is not the case for AI components, according to the head of NVIDIA. It says its Blackwell chips for data centers are more than 30 times better for AI inference than the previous generation.

NVIDIA CEO bets everything on performance

In his interview with TechCrunch, the head of NVIDIA made clear his belief that investing in performance is the only path to AI. The executive believes that this will be the best way to also reduce costs, which have been a complicated factor for many companies.

“Moore’s Law was so important in the history of computing because it reduced costs,” Huang said. “The same thing will happen with inference where we increase performance and, as a result, the cost of inference will be lower.”

Related news: