Accelerated Computing as the Core

Speaking on Dwarkesh Patel’s podcast, Huang explained that NVIDIA’s core premise is accelerated computing, combining GPUs and CUDA with CPUs to process workloads far faster than traditional methods.

NVIDIA GPUs excel at large-scale parallel processing, shifting massive amounts of code and algorithms onto GPUs to accelerate applications by 100–200 times. These applications span engineering, science, physics, data processing, and computer graphics—AI is just one of many.

“Even if AI didn’t exist today, NVIDIA would still be very, very large,” Huang said, though he admitted it would be disappointing given AI’s unprecedented role in driving revenue and growth.

China’s Computing Power

On China, Huang dismissed the notion that the country lacks AI chip capability as “completely absurd.” He argued that while the U.S. faces energy constraints, China has abundant energy resources and numerous data centers with unused capacity.

Although China cannot access advanced chipmaking equipment from ASML due to export controls, Huang noted they can compensate by producing chips in volume. “AI is a parallel computing problem. With abundant energy, you can make up for chip limitations,” he said, adding that China already dominates mainstream chip production and has long surpassed the computing thresholds many worry about.

Missed Opportunities With OpenAI and Anthropic

Huang also revealed that NVIDIA missed chances to invest in OpenAI and Anthropic when they were seeking billions in funding to scale. NVIDIA assumed such investments were best left to venture capital firms, while Microsoft, Google, and Amazon stepped in—securing commitments from those labs to use their infrastructure.

Asked about the fact that two of the world’s leading AI models, Claude and Gemini, were trained on custom chips rather than NVIDIA GPUs, Huang acknowledged the brilliance of those deals. He admitted AI would not have advanced as it has without them, but emphasized: “Next time, I won’t miss the opportunity.”

Huang’s remarks underscore NVIDIA’s broader vision: AI may be the headline driver today, but accelerated computing is the enduring foundation. With GPUs enabling exponential speedups across diverse fields, NVIDIA’s trajectory remains strong—even in a world without AI.