Jensen Huang

Who

Jensen Huang was born in Taiwan and moved to the United States as a child. He studied electrical engineering and worked at LSI Logic and AMD before co-founding NVIDIA in 1993. The company began by making graphics chips for games. Over three decades he turned it into a central supplier of computing for modern AI.

Operating rules

Invest in a platform years before the market wants it. In 2006 NVIDIA released CUDA, which let developers use graphics chips for general computing. For years it cost a great deal and drew little demand. When deep learning took off, the software and the developers were already in place.

Keep information flat. He is known for having an unusually large number of direct reports and for giving guidance in group settings rather than private meetings. Everyone hears the same reasoning at the same time, and fewer layers sit between a question and a decision.

Treat no task as beneath you. He talks openly about working in a restaurant as a teenager, including cleaning bathrooms. He tells students that resilience is built in hard seasons, not easy ones.

Think in systems, not parts. NVIDIA now sells networking, software and full data center designs, not only chips. He describes the whole data center as the unit of computing.

What I borrowed

Together with Morris Chang, he shows me two sides of one industry: the designer who bets early on a platform and the manufacturer who earns trust. From Huang I took a question I ask about my own smart building work: is this a platform other people will build on, or a one-time fix?

What I won’t copy

The language of wishing hardship on people. I believe hard work shapes character, but I would rather choose hard problems for myself than make suffering a management style.