'We'll be out of business': Jensen Huang bet his last dollars on a chip he never held in his hands — and it saved Nvidia
'We'll be out of business': Jensen Huang bet his last dollars on a chip he never held in his hands — and it saved Nvidia.
The above button links to Coinbase. Yahoo Finance is not a broker-dealer or investment adviser and does not offer securities or cryptocurrencies for sale or facilitate trading. Coinbase pays us for certain activity generated through this link. Prices displayed are informational.
Moneywise and Yahoo Finance LLC may earn commission or revenue through links in the content below.
In 1997, with his company Nvidia (NASDAQ: NVDA) mere weeks away from hitting insolvency, Jensen Huang ordered mass production of a graphics chip that had never physically existed — betting his company's last dollars on a design tested entirely inside a simulator.
The chip was the RIVA 128 and Huang called that decision, made when Nvidia had "about six months of cash left," as the defining moment that saved the company.
Jeff Bezos backs a platform that lets anyone invest in rental homes for as little as $100 — 6 ways to build wealth like a landlord without actually being one
JP Morgan sees gold hitting $6,000/oz before 2027 — and a gold IRA lets you hold the physical metal while deferring the tax bill. Get your free guide from Priority Gold
The tax breaks in Trump's 'big beautiful bill' expire after 2028 — and experts say most people won't act in time. What to do before the window closes
Huang recalled the pivotal and harrowing events during an episode of the Acquired podcast (1), recorded at Nvidia's headquarters and published back in October 2023.
At the time of that recording, Nvidia was worth about $1.1 trillion. Today, it's worth around $4.7 trillion (2), making it the most valuable public company on earth — a position it reached after becoming the first company ever to top a $5 trillion valuation (3) in October 2025.
Considering Nvidia's wild success, it's easy to forget that it was all built on a gamble, or perhaps a wish and a prayer.
To understand the risk, you have to understand what "taping out" a chip means.
In semiconductor design, taping out is essentially the point of no return: It's the moment a company sends its finished blueprint to a factory to be manufactured, or "fabricated." Fixing a mistake after that means starting a fresh production run, which costs money and months most startups truly don't have.
The normal process, Huang explained, is iterative: You build the chip, write the software, find and stamp out any bugs, then tape out a corrected version. Repeat as needed. But 30 years ago, Nvidia couldn't afford to repeat anything.
"If we only had six months and you get to tape out just one time, then obviously you're going to tape out a perfect chip," Huang said.
Huang had a darkly hilarious answer when his engineers asked how he could possibly know the chip would be perfect: "I know it's going to be perfect because if it's not we'll be out of business. So let's make it perfect."
Nvidia would use software that mimics how a physical chip will behave before any silicon exists, so Huang found a struggling company selling exactly that tool and bought its inventory right then and there, on the spot, with about half of the company's remaining cash (4).
He said his software team wrote the full stack and ran it while they sat in the lab, waiting for a single frame of Windows to render at roughly a frame a minute. But it let the team virtually prototype the entire chip, run every game and application they had and catch bugs before committing to production.
"If you're going to tape out a chip and know it's perfect, then what else would you do?" Huang said. "The answer obviously go to production and marketing blitz." So Nvidia skipped the prototype stage entirely and went to TSMC (5), the Taiwanese foundry that is now the world's largest chipmaker, for manufacturing.
