Nvidia CEO Jensen Huang Predicts ‘Hyper Moore’s Law’ Pace for AI© David Paul Morris/Bloomberg
Nvidia CEO Jensen Huang says the level of computing performance is about to grow exponentially during the coming decade, enabling new AI innovations and capabilities.
“Over the next ten years, our hope is that we could double or triple performance every year at scale,” Huang said on an episode of the AI-focused podcast “No Priors” published Thursday. “I wouldn’t be surprised, if the way people think about Moore’s Law which is 2X every couple years, we’re going to be on some kind of hyper Moore’s Law curve.”
He added: “When you double or triple every year in just a few years it adds up. It compounds really, really aggressively.”
Moore’s Law is an old forecast of innovation for the semiconductor industry by Gordon E. Moore, the co-founder of Intel. Moore said “the number of transistors incorporated in a chip will approximately double every 24 months,” offering performance and cost benefits over time.
Last year, Nvidia announced that it would begin releasing AI-data center products at a faster pace, moving from a two-year product cycle to a one-year cadence.
Huang explained that the increase in performance would enable companies to drive the cost of computing significantly lower. He also said the improvements would be driven by not only AI chip advancements but by improvements across the entire system, including software, networking, algorithms, and hardware. This would be on top of the increasing ability to scale computing workloads to larger and larger superclusters of GPUs, ranging from hundreds to thousands of units.
“The scale of the problems we could address has changed a lot,” Huang said. “We set ourselves up to be able to scale computing and develop software at a level nobody has ever imagined before.”
