A screen on the floor of the New York Stock Exchange displays news about Nvidia's stock price in New York, Monday, November 9. AP/PTI
Gaming chips to AI dominance: What Huang built, and what Nvidia now risks losing
Jensen Huang took a gaming-chip maker and made it the backbone of the AI era. Now the man at its centre is under investigation
When Nvidia’s board announced an independent investigation into Jensen Huang, it was examining allegations that he had been targeted by a foreign intelligence operation involving a woman suspected of working for China. Huang has acknowledged being unfaithful to his wife and apologised. The board’s statement said it has no current reason to believe any sensitive material was actually exposed, but that is its starting position, not a conclusion.
The stakes are high because Nvidia is Huang in a way few big companies are with their founders. He co-founded it in 1993 and has run it ever since. For years it was a niche supplier of graphics cards to PC gamers, and in 1999 it popularised the term “GPU” with the GeForce 256.
Also read: Nvidia stock slides on CEO probe while Wall Street rallies on without it
The pivotal bet came in 2006, when Nvidia released CUDA, a software platform that let developers use graphics chips for general computing. Wall Street was unimpressed, and for years the investment looked like a distraction. Then, in 2012, researchers used Nvidia GPUs to train a neural network that stunned the field at image recognition. Deep learning had found its hardware.
Trump defends Huang
From there Nvidia compounded its lead. It built chips tuned for AI, bought networking firm Mellanox to link thousands of processors together, and kept pouring money into software. Rivals could sometimes match a chip’s raw speed, but not the ecosystem. Millions of developers had learned CUDA, and the tools, libraries and habits built around it made switching costly.
When ChatGPT arrived in late 2022, Nvidia was the only company that could supply the hardware at scale. Its data centre chips became the most sought-after product in tech, and by most estimates it now holds the large majority of the AI accelerator market.
That dominance carries risks. Big customers like Google, Amazon and Microsoft are developing their own chips. Washington keeps tightening export rules on China, and analysts warn that the real fallout from the scandal could be dramatically harsher US restrictions on selling advanced chips to China.
Analysts are sceptical the controversy will force Huang out, and US President Donald Trump has defended him. Still, the timing is awkward. Nvidia reports earnings on November 17, and investors will be listening for what the company says about its future as well as its numbers.