Besides chips, Nvidia is also skilled at generating another kind of 'product'?!
This 'product' isn't hardware; it's millionaires.
A recent internal anonymous survey indicates that roughly one in every two Nvidia employees has a net worth exceeding $25 million, which translates to over one hundred million RMB.
The proportion of millionaires is even more than three-quarters.
Extrapolating from this ratio, among the colleagues you pass by daily at Nvidia's campus, over 15,000 might be wealthy enough to buy a detached house in the Bay Area outright.
A year ago, Jensen Huang, as a guest on the All-In Podcast, claimed that 'Within my management team, I've created more billionaires than any CEO in the world.'
The host immediately remarked that where NBA players once signed $300 million contracts, now it's Nvidia engineers' turn.
Those words might have sounded like bragging back then, but this survey shows that Huang's statement was actually an understatement.
The Fortune Factory with 42,000 People
This survey was first disclosed on X by The Kobeissi Letter and covered about one-tenth of Nvidia's employees.
The survey did not publish specific wealth distribution brackets but provided two key thresholds.
The first line is $1 million, with 78% of respondents above it.
The second line is $25 million, with roughly half of the respondents crossing this threshold.
The numbers for management are even more staggering.
CFO Colette Kress and Executive Vice President Jay Puri each saw their personal net worth break the $1 billion mark in July last year, becoming the latest additions to Huang's list of 'billionaires I've created.'
Moreover, in terms of scale, the number of people covered by this wealth-generating machine is equally astonishing.
As of the end of fiscal year 2026, Nvidia's global workforce totaled 42,000 employees across 38 countries and regions.
Five years ago, this number was less than 19,000, and it has increased by a net 12,400 people in just two years.
Despite this many people, the company's attrition rate is only 3.7%.
What does 3.7% mean? The average attrition rate in the US tech industry often hovers around 13%, meaning Nvidia's rate is less than one-third of the industry average.
Over 40% of new hires come from internal referrals, with existing employees bringing in new ones.
To some extent, this is understandable, as introducing a friend means helping them board a wealth express.
However, the other side of the coin isn't as glamorous.
Some former employees revealed that choosing to leave Nvidia's high-pressure work environment meant forfeiting over 15 million RMB in stock gains per year.
Leaving means abandoning a growing fortune. Staying means continuing to endure Huang's notoriously high-intensity management style.
'Golden handcuffs' — Silicon Valley has a precise name for this dilemma.
Jensen Huang himself doesn't seem to think it's a problem.
In an interview, he said he personally reviews the compensation for all 42,000 employees, guided by one principle: 'pay as much as we can.'
How the Fortune Machine Started Turning
The story of Nvidia's crazy wealth generation begins in 2008.
That year, the financial crisis swept the globe, Nvidia's stock price hit rock bottom, and the company seized the opportunity to launch its Employee Stock Purchase Plan (ESPP).
The design of this plan was clever, allowing employees to buy company stock at an 85% discount of the lowest stock price from the past two years.
What did the 2008 low price mean? It meant the cost basis for employees was so low that, in hindsight, it almost seemed like it was given away for free.
On top of the ESPP, Nvidia layered on Restricted Stock Units (RSUs) as a core component of the compensation package.
The RSU rules are standard: locked for the first year after joining, with 25% vesting in a lump sum after one year, followed by quarterly vesting batches, fully vested over four years.
A typical engineer's offer package might look like a $180,000 annual base salary plus $300,000 worth of RSUs vesting over four years.
The $180,000 base isn't particularly high for the Bay Area. The real value lies in those $300,000 worth of RSUs.
Because the value of RSUs follows the stock price.
Nvidia's stock price has surged 22,687% over the past decade, climbing from $0.87 all the way to $198.
By August this year, the stock price had further risen to around $214, with the company's market capitalization reaching $5.2 trillion, solidifying its position as number one in the world.
This means an engineer who received $300,000 worth of RSUs in 2016, as long as they never sold, would see that stock's value balloon to nearly $70 million today.
This isn't an options gamble. No one needed to bet on whether the company would go public; the stock was there from day one of employment, with time being the only variable.
ESPP plus RSUs, compounded by a stock that rose over two hundredfold in a decade — that's how this fortune machine started turning.
Good News on Top of Good News
In the same week this survey was repeatedly cited by the media, Nvidia announced another piece of news.
On August 24, at the Hot Chips 2026 conference, Nvidia announced that its dedicated inference accelerator chip, Groq 3 LPX, had officially entered full production.
This chip isn't for training, only for inference, specifically designed for AI agents.
In benchmark tests running the open-source model Gemma 4 (31B parameters, 100k token context window), the Groq 3 LPX achieved an output speed of 3,400 tokens per second, as assessed by the independent testing organization Artificial Analysis.
For latency-sensitive tasks, its response speed is 4 times faster than the closest competing platform; multi-step reasoning tasks that used to take hours are now compressed to the minute level.
The Groq 3 LPX serves as an extension module for the Vera Rubin data center platform. A single rack can integrate 256 LP30 accelerators, with cloud service provider Nebius being the first production customer.
Wealth generation relies on stock prices, and stock prices rely on products.
Nvidia has just added another link to this flywheel.
References:
[1]https://x.com/KobeissiLetter/status/1952499417681998197
[2]https://sqmagazine.co.uk/nvidia-employee-count-statistics/
[3]https://fortune.com/2026/06/02/jensen-huang-nvidia-staff-pay-ai-investment-equity/
[4]https://www.nvidia.com/en-us/benefits/money/espp/
[5]https://siliconangle.com/2026/08/24/nvidias-dedicated-inference-accelerator-groq-3-lpx-enters-full-production-to-supercharge-ai-agents/
This article is from WeChat public account 'QbitAI', author: Kexi
