Why did NVDA jump 3.7% in an afternoon session after a Reuters note about roughly ten Chinese tech firms getting cleared to purchase from the company? The short answer comes in layers. Nvidia's stock no longer trades on its own fundamentals alone; it trades on a policy weather report. Each tick of export-control news, each rumor about who can buy what from Jensen Huang's company, shows up in the price within minutes. That is what happens when a single chip designer becomes the throttle valve for the global AI buildout. The ticker is downstream of decisions made years ago in conference rooms in Santa Clara, and decisions made this quarter in Washington and Beijing. If you want to understand the volatility, the headline recap is the least useful thing you can read. The company's own history explains far more than any single news cycle will.
Most coverage of days like this stops at the mechanics. Reuters reports a licensing thaw, the stock moves, an analyst is quoted saying China exposure remains the swing factor, and the segment ends. The question of why Nvidia, specifically, sits in this position at all goes missing. AMD makes GPUs. Intel has spent decades and tens of billions trying to stay relevant in advanced compute. Broadcom, TSMC, and a dozen others occupy adjacent ground. Yet when the US government adjusts who can buy accelerators, it is Nvidia's name on the press release and Nvidia's earnings call that markets treat as a macro event. That asymmetry was built, slowly, through a sequence of strategic decisions stretching back to the early 1990s. Tae Kim's book is the most direct account of how those decisions accumulated.
Kim, a tech journalist, drew on more than a hundred interviews with Huang, his cofounders, early investors, and current executives to reconstruct the company from its 1993 founding in an East San Jose Denny's forward. The opening chapters dismantle some of the founding mythology and walk through the near-death moments that followed, including product bets that almost ended the company before it shipped anything memorable. If you have only ever encountered Nvidia as the AI stock, the early material reads like a different company entirely. The strategic core of the book is the long arc from the GPU to CUDA.
Kim treats the decision to make the GPU programmable, and then to invest heavily in a software layer most customers did not yet want, as the bet that mattered. CUDA shipped in 2006. The AI workloads that would justify it in retrospect were still years away, and for roughly a decade the investment looked like overhead on the income statement. Kim is good on how Huang's reading of Clayton Christensen's Innovator's Dilemma shaped that patience. Huang, by Kim's account, became fixated on the idea that incumbents lose by defending the wrong margins, and structured Nvidia to behave like the disruptor even after it had become large.
That meant cannibalizing existing product lines, keeping the organization unusually flat, and pushing information through direct email rather than chains of management. The organizational chapters are where the book is most concrete. Kim describes a company where Huang has dozens of direct reports, where status updates circulate as bulk emails called the Top 5, and where engineers can escalate past their managers without political cost. Whether this scales gracefully past current headcount is a fair question the book does not fully answer. Flat structures tend to depend on a single person's bandwidth, and Huang is sixty-two.
On the AI turn itself, Kim makes a stronger case than I expected. The standard story is that Nvidia got lucky when deep learning researchers discovered GPUs were good at matrix math. Kim shows the company actively courting those researchers years before ChatGPT, sending engineers to academic labs, subsidizing tooling, and treating a tiny customer segment as a strategic priority. Luck is the wrong word for a decade of deliberate seeding. The book is weaker on the present. Kim is admiring of Huang in a way that softens the harder questions about concentration risk, the political economy of export controls, and what happens to a company whose addressable market is now partly set by State Department memos. The history is excellent. The forecast is too gentle on a subject whose risks deserve sharper handling.
So back to the opening question: why did the stock move on a narrow licensing item? Because when one company sits at the center of the AI supply chain, narrow licensing items are the story. The Nvidia Way is a fair place to spend a weekend if you want the decisions behind the ticker rather than another recap of the ticker itself. You will not come away with a price target. You may come away knowing which future headlines are worth your attention, and which ones are just weather.