Nvidia's AI Chip Breaks Moore's Law

According to TechSpot , at the CES 2025 exhibition (held in Las Vegas, USA), Nvidia CEO Jensen Huang made a shocking statement: the company's AI (artificial intelligence) chips are progressing faster than Moore's law.

CEO Jensen Huang claims Nvidia's AI chip has surpassed Moore's law

PHOTO: TECHSPOT SCREENSHOT

Nvidia is breaking all the limits

Moore's Law has been the guiding principle for the chip industry for decades, predicting that the number of transistors on a chip will double every year. However, in recent years, this law has shown signs of slowing down.

Nvidia’s AI chips are now 1,000 times more powerful than they were a decade ago, a growth rate that far exceeds Moore’s Law predictions, Huang said. He attributed this to Nvidia’s holistic approach, from architectural design, chip manufacturing, system building, to software and algorithm development.

The clearest proof of Mr. Huang's statement is Nvidia's latest data center super chip GB200 NVL72. With AI task processing performance 30-40 times faster than the previous generation, GB200 NVL72 is expected to help reduce the cost of running complex AI models, opening the door to more advanced AI applications.

CEO Huang also dismissed the notion that AI development is slowing down, emphasizing the importance of increasing ‘test-time’ capabilities, which give AI models more time to ‘think’ and come up with more accurate answers.

With breakthroughs in AI chip technology, Nvidia is consolidating its lead in the global AI race. Tech giants such as Google, OpenAI, and Anthropic are all relying on the company's chips to develop advanced AI models. Mr. Huang's statement shows that Nvidia is not only keeping up but also setting new standards for performance and cost-effectiveness for AI chips, contributing to the acceleration of the AI ​​revolution worldwide .

Source: https://thanhnien.vn/chip-ai-cua-nvidia-pha-vo-dinh-luat-moore-185250109133650152.htm