Huang Renxun: The performance improvement speed of AI chips has surpassed Moore's Law

PANews reported on January 8th that according to TechCrunch, NVIDIA CEO Huang Renxun stated in his CES speech that the performance improvement speed of the company’s AI chips has surpassed Moore’s Law. He pointed out that NVIDIA’s latest data center super chip GB200 NVL72 is 30-40 times faster than the previous generation H100 in AI inference tasks, and the overall chip performance has increased by 1000 times compared to ten years ago. Huang Renxun emphasized that by synchronously innovating in architecture, chips, systems, libraries, and algorithms, NVIDIA can achieve progress beyond Moore’s Law. He also predicted that as chip performance continues to improve, the cost of high-computing-demand AI inference models such as OpenAI’s o3 model will gradually decrease.