Name: The Thinking Machine: Jensen Huang, Nvidia, and the World's Most Coveted Microchip
Rating: 4 (3 reviews)
ISBN: 0593832698

Stephen Witt

The Thinking Machine: Jensen Huang, Nvidia, and the World's Most Coveted Microchip

Jose Sbuck

This book tells the story of Jensen Huang, an electrical engineer who founded Nvidia and has remained its CEO for more than 30 years. Nvidia created the GPU category: it started in computer graphics and eventually became central to deep learning and AI. The story is fascinating, full of close calls, bold bets, visionary decisions, and a fair amount of luck. Huang himself is busy, successful, and somewhat of an enigma. The author interviews Huang, his friends, and his colleagues, but reveals little beyond the standard anecdotes about an intelligent, hard‑working, enthusiastic, driven person. This is very much in line with other successful entrepreneurs. Whatever truly differentiates him from the rest of us remains largely unexplored. We’re repeatedly told he is “hard working” and obviously bright, but the deeper “why” of his success never really surfaces. Nvidia as a company is particularly interesting. I’ve long seen it as something of a bad actor: always pushing for customer lock-in, building proprietary solutions that don’t play nicely with the open-source ecosystem. In my mind, it’s the stereotypically greedy company that tries to extract ever more money from its customers while hindering competition and broader industry development. The book paints a very different picture. According to Witt, Nvidia drove the GPU revolution and invested heavily in CUDA (and cuDNN), making them available in ways that allowed scientists and others to harness unprecedented compute for new kinds of products. The book also highlights Nvidia’s ability to iterate new, more efficient chips at a pace that far exceeded industry norms. They relied heavily on simulation, took big risks, and worked long hours to pull it off. Witt also introduces some of Jensen Huang’s more distinctive concepts. One is Nvidia’s search for the “speed of light” in its operations. It is the absolute fastest conceivable way to get something done if everything goes perfectly and there are no obstacles. That “speed of light” becomes the target. If Nvidia can reach it in any particular process, they can be confident no competitor will move faster. It’s a neat encapsulation of Huang’s philosophy: stay alive by being better and faster than everyone else. Another is the idea of the “zero billion dollar market”: essentially a blue ocean, a market that doesn’t yet exist but can be created through innovation. Ten years ago, there was no demand for AI GPUs. By building unprecedented computational capacity and a software layer to exploit it, Nvidia effectively created its own category and its own market. A market, where it remains the dominant, and for a long time the only, serious vendor. Of course, this was extremely risky and only worked because researchers happened to discover that deep learning runs very well on GPUs and that neural networks could outperform previous approaches. <blockquote> A well-trained model mimicked (and sometimes exceded) our own capacity for cogitation, but it had no sense memory, no emotions, no imagination, no reproductive function, and no instinct to survive. In this sense, one could not say it was really alive-it did not have the self-sustainability of even the most primitive single-celled organism. Although there was certainly an analogy between training neural nets and evolution, AI was not the product of the kind of ruthless, Darwinian, kill-or-be-killed, eat-or-be-eaten planetary laboratory from which biological life had emerged. </blockquote> These ideas made me want to finally pick up Christensen’s <a href="https://www.goodreads.com/book/show/2615.The_Innovator_s_Dilemma_The_Revolutionary_Book_that_Will_Change_the_Way_You_Do_Business" title="The Innovator's Dilemma The Revolutionary Book that Will Change the Way You Do Business by Clayton M. Christensen" rel="nofollow noopener">The Innovator's Dilemma: The Revolutionary Book that Will Change the Way You Do Business</a>, which the book references multiple times. They also bring to mind Risto Siilasmaa’s <a href="https://www.goodreads.com/book/show/42585317.Paranoidi_optimisti" title="Paranoidi optimisti by Risto Siilasmaa" rel="nofollow noopener">Paranoidi optimisti</a>. The common thread: staying in business over the long term requires staying genuinely innovative. As engaging as the story and subject matter are, the book itself feels a bit rushed. The narrative often reads like an extended magazine feature stitched into book form, with a tendency to underplay the broader team effort. Witt occasionally singles out individuals and then spends time on their appearance, clothing, or tone of voice, the usual journalistic tricks for making a story feel “relatable.” He also drifts into first-person speculation without making a strong case for why his musings deserve special attention. At times, the narrative even wanders away from Nvidia and Huang into potted histories of deep learning that feel generic and boilerplate, almost like they were generated by an AI text model. Of course, writing a definitive, reflective account is difficult when the field is evolving so rapidly. Perhaps this book is best read as timely reporting rather than deep analysis. Recommended for readers interested in chip design, GPUs, and the hardware layer behind modern AI systems.

The Thinking Machine: Jensen Huang, Nvidia, and the World's Most Coveted Microchip

Reviews

Highlights