Nvidia CEO Jensen Huang Claims AI Chips Surpassing Moore’s Law

Nvidia CEO Jensen Huang Claims AI Chips Surpassing Moore’s Law

Nvidia CEO Jensen Huang declares AI chip performance is accelerating beyond Moore’s Law, reshaping AI capabilities with groundbreaking advancements in Nvidia’s chip technology.

Jensen Huang, CEO of Nvidia, has boldly said that the company’s AI chips are getting better faster than Moore’s Law allows. His words bring attention to Nvidia’s ongoing innovations, which could change how quickly and cheaply AI is developed.

In 1965, Moore’s Law said that every two years, the number of transistors on a chip would double, which would double its computer power. In the past few years, this rate has slowed down, but Nvidia’s AI chips are speeding it up. Huang says that their newest super chip is more than 30 times faster than the one that came before it at AI inference jobs.

Huang says that Nvidia’s fast progress is due to the fact that they are constantly coming up with new designs, chips, systems, libraries, and algorithms. He told TechCrunch, “This integration lets us move faster than Moore’s Law.”

A lot of the best AI labs, like OpenAI, Google, and Anthropic, use Nvidia chips to train and run their models. Huang says that the faster performance of Nvidia’s chips directly leads to AI systems that are smarter and cost less.

It costs a lot to run AI models that use test-time math, like OpenAI’s o3. Huang thinks that Nvidia’s cutting-edge GB200 NVL72 chip—which is 30–40 times faster than its H100 chip—will greatly lower these costs, allowing more people to use advanced AI.

Huang thinks that AI chips will not only make things work better, but they will also make them cheaper. “As inference performance goes up, the cost of AI inference will go down,” he said, stressing that future AI reasoning models will be affordable.

In the past ten years, Nvidia tripled the bars, which means it made chips 1,000 times faster than it should according to Moore’s law. Huang is confident that this trend will persist, and will improve the AI apps.

Nvidia’s Jensen Huang is now plotting to end the speed at which AI hardware is being produced. Here, the company is pointing the way to making new AI solutions stronger, more efficient, and cheaper than has previously been possible. These solutions could alter the stability of various businesses across the globe.

Leave a Reply

Your email address will not be published. Required fields are marked *

Elon Musk Warns AI Training Data is Exhausted, Calls Synthetic Data Key Previous post Elon Musk Warns AI Training Data is Exhausted, Calls Synthetic Data Key