The Mindset That Built NVIDIA

NVIDIA began by trying to reinvent 3D graphics for PCs, but Huang says its original algorithm and technology were wrong. In 1995, he confronted the problem with the team and bought three textbooks on OpenGL and graphics pipelines. He says the company’s lasting lesson was to face reality, learn what was missing, and tackle important problems even when they were new to the team.

Huang says NVIDIA’s central idea was to augment CPUs to solve problems that were otherwise difficult, and to accelerate algorithmic domains rather than focus only on chips. He names molecular dynamics, image processing, inverse physics, and deep learning as examples. When AlexNet appeared, NVIDIA investigated deep learning as a general approach that could affect software, computer vision, robotics, and self-driving cars.

Huang describes curiosity as the reason he studies technical details: understanding a fast-changing field helps him inform himself, serve the company, and share useful insights. He compares leading to learning to read a wave, which requires getting close enough to understand what is happening. He also says founders should shape the organization around how they work, and that the next CEO can reshape it later.

Huang argues that systems design will matter as agents automate more low-level chip and software work. He says people need to reason about inputs, outputs, information flow, and constraints such as processors, memory, and networking. For robotics, he describes training through simulation and reinforcement learning, and estimates physical AI could become a $100 billion business for NVIDIA in more than two but fewer than ten years.

Huang advises students to learn hard sciences, engineering, and systems thinking as routine coding becomes more automated. He says AI can help people take on larger problems, while identifying unmet needs and understanding how technology intersects with society remain valuable. His advice to founders is to trust their ability to learn, start despite uncertainty, and build resilience by getting through each day.