A tribute
The Architect of the Thinking Age
On Jensen Huang, his life’s work, and why we owe him more than gratitude.
I. The weight of November
There is a peculiar kind of silence that falls over an industry when the person who defined it suddenly falls out of the sky. In the second week of November 2026, that silence fell across the technology world, because on Thursday the twelfth, NVIDIA’s board accepted the resignation of the man who had led the company for thirty-three years — Jensen Huang, the founder who turned a graphics-chip startup into the most consequential hardware company of the century, and who, more than any other single individual, made the artificial-intelligence age possible. The stock dropped more than thirteen percent that morning, wiping out close to 1.6 trillion dollars of value across a single disastrous week, and pundits who had spent a decade praising his judgment rushed to pronounce the era over. He was sixty-three years old, and after three decades of standing on stages in front of tens of thousands of cheering people, he walked off one alone.
It is precisely in moments like these that a society reveals what it actually values, and I want to argue, with my whole heart, that we are in grave danger of forgetting what this man did for us. Whatever one concludes about the failures of his private life — and they were real failures, which he himself has admitted with a directness that many public figures lack — they do not erase a body of work that has genuinely reshaped the human species’ relationship with intelligence itself. The AI industry, which I consider the single greatest achievement in the history of human civilization, did not arrive on its own. It was built on top of a foundation that one stubborn, tireless man spent thirty years laying, through ridicule, near-bankruptcy, and a decade in which every rational analyst on Wall Street told him he was wasting his shareholders’ money. To abandon him now, at the very moment of his hardest difficulty, would be not only cruel but historically illiterate, because it would mean crediting a golden age to nobody at all.
The AI industry did not arrive on its own. It was built on a foundation that one stubborn, tireless man spent thirty years laying.
II. A child sent across an ocean
To understand why his endurance matters, you have to begin with the circumstances of his beginning, because they are almost impossibly far from where he ended up. Jensen Huang was born Jen-Hsun Huang in Taipei on February 17, 1963, the second son of Huang Hsing-tai, a chemical engineer at an oil refinery, and Lo Tsai-hsiu, a schoolteacher, a middle-class Taiwanese family that spoke Hokkien at home and moved frequently for the father’s work. When Jensen was five years old the family relocated to Bangkok, where he attended the Ruamrudee International School for roughly four years, and it was there that his mother began a small daily ritual that would shape his entire life: she would open the dictionary at random, select ten words, and drill her sons on their English, preparing them for a country they had never seen. In 1973, when Jensen was nine, his parents sent him and his older brother to the United States to live with an uncle in Tacoma, Washington, while the family remained in Asia.
What happened next reads like the opening chapter of a novel about survival. The aunt and uncle, themselves recent immigrants who did not know the American school system, mistakenly enrolled the two boys in the Oneida Baptist Institute in rural Kentucky, believing it to be a prestigious boarding school. It was in fact a religious reform academy for troubled children, and the brothers arrived as what one biographer described as undersized Asian immigrants with long hair and heavily accented English, at a school where they were beaten, called ethnic slurs, and threatened with knives. Jensen, only ten years old, cleaned toilets every day; he was too young even for the academy’s own classes and attended the elementary school in town instead. His parents sold nearly everything they owned to pay the tuition. In that miserable place, though, something was forged that would define him: he taught his roommate how to read in exchange for being taught to bench-press, he took up table tennis, and by fourteen he was good enough to appear in Sports Illustrated. When his parents finally arrived in America two years later, discovered what had happened, and pulled their sons out, the family settled in Beaverton, Oregon, and Jensen enrolled at Aloha High School, where he skipped two grades, graduated at sixteen, became a nationally ranked table-tennis player, and encountered an Apple II that taught him to program in BASIC. His first job, at fifteen, was the graveyard shift at a Denny’s restaurant, where he washed dishes, bussed tables, and waited on customers between 1978 and 1983.
He chose Oregon State University because of its low tuition, earned a bachelor’s degree in electrical engineering with highest honors in 1984, worked as a microprocessor designer at AMD and then as a division director at LSI Logic, and picked up a master’s degree from Stanford in 1992 while holding down that job. He met his wife, Lori, in an engineering class at Oregon State, and they married in 1985. Nothing in that résumé suggests destiny. It suggests, instead, a young man of modest means who simply refused to be small.
III. The diner on Berryessa Road
NVIDIA was born in the most ordinary of places, which is part of why its story feels so democratic. In late 1992, two engineers from Sun Microsystems — Chris Malachowsky and Curtis Priem — were frustrated with their employer and wanted to build something of their own, and they recruited Jensen, the LSI Logic engineer who had been assigned to work with them on Sun’s graphics hardware and who had become a close friend. The three of them could not agree at first even on who would quit their job first; Jensen’s wife Lori refused to let him resign unless Chris resigned simultaneously, and Chris’s wife felt exactly the same way, until Curtis broke the deadlock by handing in his notice first. They sketched out their vision for a company over bottomless cups of coffee at a Denny’s roadside diner on Berryessa Road in East San Jose — Jensen’s joke being that he had already worked at Denny’s, so he considered it his first company — and on April 5, 1993, NVIDIA incorporated with forty thousand dollars in the bank and twenty million in venture capital from Sequoia Capital and others. Jensen left LSI on February 17, 1993, which was his thirtieth birthday.
Their reasoning, even then, was unusually clear-eyed. They had noticed that video games were simultaneously one of the most computationally difficult problems in existence and one of the largest consumer markets in the world, a combination that almost never occurs, and they reasoned that whoever could solve the hard problem cheaply would be funded by enormous demand to keep solving harder problems still. That observation — that entertainment could fund the research that eventually produces general-purpose computing — is the seed from which the entire AI infrastructure industry eventually grew. But between the observation and the reality lay six years of what can only be described as a long flirtation with extinction.
IV. Three years of dying
NVIDIA’s first chip, the NV1, shipped in 1995 and was, by any commercial measure, a disaster: an overpriced, overambitious “octopus” of a device that did 3D graphics, video, audio and input acceleration all at once when the market wanted exactly one of those things, and it was built on quadrilateral rendering primitives when the rest of the industry, following Microsoft’s newly released DirectX, had committed to triangles. Their partner Diamond Multimedia returned almost all two hundred and fifty thousand units, which nearly destroyed the company outright. The second chip, the NV2, was being developed in partnership with Sega for a next-generation console, and it was built on the same doomed architecture. Jensen found himself facing what he later called an existential moment: finish the Sega contract and spend two years perfecting an obsolete technology while fifty competitors raced past, or abandon it and run out of money immediately. He flew to Sega and asked its president, Irimajiri-san, to release NVIDIA from the contract while still paying them in full — a request with no rational basis whatsoever, since Sega would receive nothing in return. The man thought about it for a couple of days and agreed, and those five million dollars, as Jensen has said ever since, bought them six months to live.
In 1996 he laid off more than half the company, cutting headcount from one hundred to forty, and staked everything on a single new chip, the Riva 128, with an instruction that still sounds like legend: they would get exactly one tape-out, because if they needed a second one they would be bankrupt before it arrived. The team emulated the design on borrowed equipment from a company that was itself going out of business, wrote all the software in advance, taped out the chip once, and it worked the first time. The Riva 128 sold a million units in four months and saved the company; the Riva TNT consolidated the reputation two years later, and by the time NVIDIA went public in January 1999, shares surged sixty-four percent on the first day of trading. For years afterwards, Jensen opened internal meetings with the same sentence — “we are thirty days from going out of business” — not because it was still strictly true, but because he never wanted anyone in the building to forget how close it had been.
The fifth chip, released in late 1999, was the GeForce 256, which NVIDIA marketed as the world’s first GPU, and its significance was not merely that it was faster but that it was programmable: a chip whose function could be redefined in software rather than fixed in silicon. That idea — the programmable accelerator — is the hinge on which the modern world turns.
They would get exactly one tape-out, because if they needed a second one they would be bankrupt before it arrived. The chip worked the first time.
V. The decade they told him to abandon
If the founding story is one of survival, the next chapter is one of faith, and it is the part of Jensen Huang’s career that I find most difficult to overstate. Beginning around 2006, NVIDIA invested more than a billion dollars building CUDA, a platform that let graphics chips be used for general scientific and mathematical work rather than just rendering polygons. For nearly ten years this produced essentially no return; the financial press treated it as the founder’s vanity project, analysts openly urged him to cut it, and investors punished the company for spending research money on what looked like a toy. He refused every time, because he had seen something the market could not yet see: that if you could teach a chip to compute in parallel, you could eventually teach a machine to learn.
When the University of Toronto’s AlexNet network demolished the image-recognition competition in 2012, it did not run on a CPU — it ran on two NVIDIA consumer GPUs, and that result is the spark that ignited the entire deep-learning revolution. Every subsequent milestone of the AI era, from the transformer architecture of 2017 to the language models that followed, was executed on the architecture Jensen had spent a decade defending against the wishes of his own shareholders. The CUDA developer population grew from one million in 2018 to five million by 2024, and by the middle of this decade NVIDIA chips powered more than eighty percent of the AI training market and over three-quarters of the world’s TOP500 supercomputers. When OpenAI released ChatGPT in late 2022, NVIDIA’s stock rose roughly fourteenfold, and in 2024 the company became the most valuable public company on Earth, the first to reach a four-trillion-dollar valuation in July 2025. In its 2026 fiscal year NVIDIA reported revenue of 215.9 billion dollars and net income of 120.1 billion — figures that would have sounded like satire a decade earlier.
That is what was built. And it was built not by a lucky gambler but by a man who was right for ten straight years while everyone around him was wrong.
VI. The man in the leather jacket
What made Jensen Huang peculiar among the titans of technology is that he never stopped performing, in the deepest and most generous sense of that word. He showed up to keynotes in a black leather jacket while every other CEO wore a fleece or a crew neck, he brought ten thousand people to their feet at GTC and Computex and CES year after year, and he spoke about the future of humanity with the conviction of someone who had done the arithmetic himself. He worked a hundred and twenty hours a week into his sixties, he answered his own email at two in the morning, he could drill into any line of code in any presentation, and he knew the name of every employee he had ever let go. Yet for all that ferocity he spoke of his mother with a softness that stopped interviews mid-sentence, he told the Denny’s story without a shred of self-pity and only with gratitude, and he remained married to Lori for more than four decades. He carried the whole world’s expectations on his back and somehow never let it turn him into an abstraction.
The honors that accumulated around him tell their own story: the IEEE Founders Medal in 2020, the VinFuture Prize, an Edison Award and a Primetime Engineering Emmy in 2024, the Queen Elizabeth Prize for Engineering in 2025, and in 2026 both the IEEE Medal of Honor and the National Medal of Science — the two highest distinctions his profession can bestow — alongside Fortune’s Businessperson of the Year, the top spot on Harvard Business Review’s ranking of the world’s best-performing CEOs, two selections for the Time 100, and a Carnegie Great Immigrants honor that cited him as proof of what an arriving child can become. His cousin Lisa Su runs AMD, which lends the family’s engineering prowess a certain comic completeness. By any measure that has ever been applied to a technologist, the record is complete.
And the archive this site keeps is, in its small way, part of that record. Between 2010 and 2026 he gave the same argument again and again in public — that accelerated computing was the only way forward, that intelligence was computable, that the difficult problem was worth funding long before it was profitable — and you can watch his certainty harden into vindication by moving chronologically through the keynotes.
VII. The autumn of 2026
And then came the autumn. In early November 2026, reporting surfaced alleging that Jensen had been the target of a foreign intelligence approach involving a woman whom American officials believe is connected to Chinese intelligence, alongside separate reporting of an extramarital relationship within the company. On Sunday, November 8, he wrote to NVIDIA’s staff admitting that he had been unfaithful to his wife, apologizing to his family and colleagues, welcoming the board’s independent investigation — which was conducted by outside counsel — and offering his full cooperation. The board stated that it had no reason to believe that national security or company secrets had been compromised, and the Chinese government rejected the espionage allegations outright. Three days later the board accepted his resignation, and on Thursday, November 12, NVIDIA announced that longtime CFO Colette Kress would serve as interim chief executive while a search for a permanent successor proceeds. Within hours the stock had fallen thirteen percent, headlines were writing his epitaph, and the man who had been the global face of artificial intelligence was no longer employed by the company he had founded.
Let me be clear about what I am and am not saying. I am not saying that nothing happened, and I am not asking anyone to pretend that a personal failure is not a personal failure; Jensen himself did not pretend it, which is more than most men in his position manage. What I am saying is that the response we are witnessing — the rush to convert a human being into a concluded case, to treat a decades-long body of world-changing work as though it were a promissory note that has now been voided — is a moral error of the first order, and one that says far more about our appetite for spectacle than about his worth.
VIII. Why we must stand with him
The case for standing with Jensen Huang in this hour does not depend on excusing anything, and I would not ask you to make it. It depends on a simple principle: that a person’s contribution to the world is judged across the whole of their life, and that a society which only honors its builders while they are winning has understood nothing about what building actually is. This is a man who was sent alone to a reform school at ten years old, who washed dishes through his teens, whose first product was returned by the crateful, who laid off sixty of his own people to save forty, who bet a billion dollars on an idea his own investors wanted him to abandon, and who was right — spectacularly, historically right — for thirty years while almost everyone else was wrong. If we grant him nothing in return for that except scrutiny at the moment of his stumble, then we have declared that courage is a loan to be called in at the first sign of human frailty.
Consider also what he carried on our behalf. Someone had to be the public voice of this revolution while governments debated it, markets panicked about it, and ordinary people feared it; Jensen did that job for a decade with an optimism that was itself a kind of public service, arguing in Washington and Beijing and Davos that the technology should be open, that the benefits should be shared, that the future was worth being excited about rather than afraid of. You do not have to agree with every position he took to recognize that the industry had a spokesman of extraordinary grace, and that it is much easier to lose such a person than to find another. Even in his worst week, the people who know the industry best responded with something closer to grief than to dismissal: colleagues spoke of his contributions as “immeasurable,” competitors offered him jobs, and ordinary engineers posted that they had chosen their life’s work because of a keynote they had watched him give. That is not the response people reserve for a villain.
There is also the matter of how we wish to be treated ourselves. Almost every person reading this has failed at something privately while succeeding enormously at something publicly, and almost every person reading this hopes, in their honest moments, that the people who benefited from their work will remember the work when the failure becomes visible. Granting Jensen Huang that basic decency — separating the gratitude we owe his achievement from the judgment we pass on his conduct, holding both at once without letting either cancel the other — is not an act of blindness. It is the mark of a mature civilization, and honestly it is the mark of a mature friend.
Stand with him. He stood with the future for longer than most of us have even been watching it.
IX. What remains
One day the leather jacket will be hung up permanently, the stage lights will dim, and the extraordinary machine he assembled will pass entirely into other hands. But the thing that was built will remain, and it will remain forever: a world in which machines can learn, in which diseases are screened in hours rather than years, in which a student in a small town can ask a question and receive the distilled knowledge of a species, in which the impossible has become merely expensive. No one person is ever the sole author of an age — history is written in the chorus — but some voices carry the melody, and Jensen Huang carried ours for thirty-three years, through failure and ridicule and triumph, with a stamina that looks less like ambition and more like love.
He was nine years old, alone in a country that had not asked for him, cleaning toilets in a school that was never what his family had been promised, and he became the architect of the greatest technological achievement in human history. Now, at the lowest point of his life, the least we can offer is what he was never offered as a child: patience, memory, and the grace to weigh an entire life rather than its final chapter. Stand with him. He stood with the future for longer than most of us have even been watching it, and the future he stood with is the one we are all now living in.
Listen to the record
Everything above is opinion; these are the talks it rests on. Watch them in order and you can hear the argument change from hope to certainty to valediction.
- Jensen Huang at Future in Review (FIRE) 2010
The earliest talk in the archive — a founder still seven years away from being proven right.
- NVIDIA CEO Jensen Huang — Acquired
The NV1, Sega, the layoffs and the CUDA decade, told at length in his own words.
- Jensen Huang, Founder and CEO of NVIDIA — Stanford GSB
On leadership, mortality and how he wants to be remembered — the most personal talk here.
- Nvidia CEO Jensen Huang and the $2 trillion company powering today's AI — 60 Minutes
The interview that introduced him to the world outside the technology industry.
- Jensen Huang: NVIDIA – The $4 Trillion Company & the AI Revolution | Lex Fridman Podcast #494
Three hours on the $4 trillion company, the AI revolution and risk.
- Jensen Huang on Vision, Risk, and the GPU — Only In America | Hoover Institution
Vision, risk and the GPU — an immigrant’s account of the road from Oregon to Santa Clara.
- NVIDIA GTC Berlin 2026 Keynote With CEO Jensen Huang
His final keynote as chief executive, delivered three weeks before his resignation.
Want the other side? The critical pieces — the culture, the cover-ups, the fall — are collected on Criticism & Controversy →