Stephen Witt’s book tells the story of how Jensen Huang and Nvidia became leaders in artificial intelligence. Witt highlights key moments that sparked this revolution, including the invention of SuperVision, a neural network that could identify objects in images, and the development of parallel processing by researchers like Bryan Catanzaro at UC Berkeley. These breakthroughs, however, faced resistance from the scientific community, which initially dismissed the potential of AI and parallel processing. It was Jensen Huang’s vision that saw the potential of these technologies and helped Nvidia conquer the field.
Stephen Witt did not write the book of Nvidia, and maybe not even the one on Jensen Huang—it’s doubtful anyone could write the book of Jensen Huang. But Witt did write the book on the synergistic collision of Jensen Huang and AI, and he got it all in a page-turning story with never-before-heard antidotes and insights.
Witt traces the eureka moments when University of Toronto Professor Geoffrey Hinton assigned Alex Krizhevsky and Ilya Sutskever, two Russians, to work on neural net design. Trying to replicate a neuron, they realized a powerful parallel processor like Nvidia’s gaming GPU was what was needed—it truly was the catalytic moment that created the path for Nvidia to take over the world.
The second big breakthrough was when Fei-Fei Li at Princeton and later Stanford launched the ImageNet challenge to see if a computer could successfully identify features, characters, and objects in images. That turned out to be the perfect test for Krizhevsky and Sutskevers’ neural net, which they had named SuperVision, and it showed off the power of a GPU.
Almost at the same time, Bryan Catanzaro at the University of Berkeley was convinced parallel processing was the only way Moore’s law, which he considered dead in the mid-2000s, could be realized. He started the UC Berkeley Parallel Computing Lab and concluded 3D graphics was the killer app.
Despite their vision, breakthroughs, and demonstrations, like Curie and other scientists before them, the AI and parallel professional scientists found that the established scientific and computer scientists dismissed them and said there was no future in their work, which meant no funding.
Jensen, however, was looking for ways to exploit the parallel processing power of his GPUs beyond just games, and he did see it. He saw it vividly and viscerally, and suddenly, a vision formed.
The book combines the history of Huang’s life with the history of the development of AI as we know it today.
Huang is like no other CEO in the computer industry. The book captures his intensity, dedication, and drive—for himself and his employees.
Tracing the development of AI through Witt’s book, I found myself getting excited; it read like an adventure novel, and I could easily imagine it as a movie. I suspect it will become one.
Witt has made a tremendous investment in writing this book over a few years. He interviewed dozens and dozens of people, held meetings with Huang, and got yelled at by Huang (many consider that an honor). Not many people would have the stamina and courage to take on such a project. It’s a monument and will take its place in the history of computers, Silicon Valley, and, most certainly, AI. Huang may not read it; he hates talking about and reading about himself, and Witt found that out too.
You can find the book here.
LIKE WHAT YOU’RE READING? INTRODUCE US TO YOUR FRIENDS AND COLLEAGUES.