Near-death
Two brushes with extinction in four years — and the mantra they left behind.
Death #1: The NV1 (1995–1996)
NVIDIA’s first chip, the NV1, was a bet on quadratic texture mapping — an elegant approach to 3D graphics that was incompatible with the triangle-based standard Microsoft’s DirectX was establishing. The market chose triangles. The NV1 flopped.
For a startup that had raised venture money on the promise of graphics leadership, a failed first product is usually fatal. NVIDIA’s investors — Sequoia Capital and Sutter Hill — watched their bet go sideways within two years of founding.
The Sega lifeline — and its collapse (1996–1997)
NVIDIA’s bridge to survival was a contract with Sega to build the graphics chip for its next game console. But the same standards problem doomed it: NVIDIA couldn’t deliver a DirectX-compatible part on Sega’s schedule.
Huang’s telling of what happened next has become Silicon Valley lore. He went to Sega, told them plainly that NVIDIA couldn’t fulfill the contract, and asked to be released. Sega’s CEO agreed — and, in Huang’s retelling, Sega paid NVIDIA around $5 million for the work completed, effectively a parting gift that funded the company’s survival.
Two honesty notes, both on The myth machine: the release is the documented core; the dollar figure is reported, not filed. And the story’s moral — radical honesty with partners — is Huang’s framing, though Sega’s generosity is not disputed.
The layoffs
Survival had a price. NVIDIA cut deep — the company shrank dramatically, with only a core team remaining to bet everything on one last chip. Huang has described this period as the hardest of his career: laying off people who had believed in the vision, then asking the survivors to work as if the company had a future.
Death #2 averted: RIVA 128 (1997)
The bet-everything chip was the RIVA 128, built this time for the DirectX triangle standard. It shipped in August 1997 — and reportedly sold over a million units in its first four months, generating the revenue that funded the next generation.
The company’s bank balance at the low point reportedly covered about a month of payroll. From this comes the mantra Huang still quotes: “Our company is thirty days from going out of business.” A slogan, not a filing — but one he says he kept repeating deliberately, long after the danger passed, to preserve the paranoia.
Why he keeps telling it
Huang returns to the near-death years constantly — in keynotes, in the Stanford commencement address, in interviews. The stories do triple duty:
- Credibility. He earned the right to preach resilience; he has the scars.
- Culture. “Thirty days” is a management tool: a profitable, dominant company that feels near death doesn’t get complacent.
- Permission to bet. CUDA (2007) was greenlit by a CEO who had already survived being wrong. The near-death years are the psychological foundation of the CUDA bet.
The counter-reading
A skeptic could note: surviving near-death twice is also luck — Sega’s generosity, the timing of RIVA 128, investors who didn’t pull the plug. Huang’s telling emphasizes character and honesty; the record also shows fortunate timing and forgiving partners. Both can be true.
Source notes
The NV1/RIVA 128 history draws on NVIDIA’s corporate records, contemporaneous press coverage, and Huang’s interviews including the Sequoia oral history. The Sega payment figure, RIVA 128 unit sales, and payroll-runway details are reported in profiles and retellings, not confirmed in filings, and are flagged throughout.