In its early days the company focused on video games, developing components that allowed games consoles and personal computers to render three-dimensional graphics by stacking countless microscopic triangles on top of each other. It went public during the dotcom bubble at a $US625 million valuation.
While the company became a key player in the gaming industry, it would be another decade before the seeds of its present success were sown. In 2012, scientists at the University of Toronto, led by the British computer scientist Geoff Hinton, developed an image recognition program that used Nvidia’s graphics chips, rather than the central processing units that most computers relied on.
Hinton would later describe it as the “big bang moment” for today’s AI movement. The program, built on two Nvidia chips bought on Amazon, blew away the competition in an annual machine vision competition and its underlying architecture spawned the deep learning craze.
Nvidia became the world’s most valuable company last week. Credit: Bloomberg
Nvidia’s chips being perfect for machine vision was something of a happy accident, but Huang capitalised. While AI was still a minor academic pursuit compared to the lucrative video games business, Huang told managers that Nvidia would become an AI business and threw resources at developing advanced chips that could power synthetic intelligence.
“It turned out to be really useful for stuff we didn’t know it was going to be useful for,” says a former Nvidia executive. “But even then you could see that [Huang] bought into the idea of a use of the GPU outside of graphics and gaming.
“He had a vision at that point that it would do something, he probably didn’t know what, but he could see there was an opportunity that a lot of us didn’t see.”
“He’s building a church, and has been laying it brick by brick for years,” says Hsu. “One day all of a sudden you see this magnificent church in front of you, but he’s been toiling for years to make it perfect.”
The rewards have now become obvious. Last month Nvidia said that revenues had risen by 262 per cent in the last quarter, to $US26 billion. Profits have climbed more than sevenfold, from $US2 billion to $US15 billion. The waiting list for the company’s superpowered chips runs into next year.
Huang has spent years nurturing relationships with major customers and suppliers, hand delivering initial orders. At the same time, he has tied users into a software system that has given the company an unassailable lead, despite rivals seeking a piece of its fortune.
Despite his approachable public persona, within Nvidia he is known as a demanding boss. Huang makes a point of quizzing even junior employees about their work and can dash off hundreds of emails to executives in a day.
“He’s not the right guy to bullshit,” says a former employee. “The people who can work in Jensen’s world have no hiding place for success or failure. The people he struggles with are the people who typically don’t respond in the right way towards mistakes.”
Loading
“Any weakness, real or imagined, is called out and people can be publicly beaten up,” says another former executive. One of Huang’s own missteps came in 2020 when he agreed to pay $US40 billion for the British microchip company Arm. After almost two years, regulators kiboshed the deal, a fact Huang still laments.
Matching his standards can be exerting. Huang is known to wake as early as 4am (“As a CEO, not sleeping is a good choice,” he once said). Huang told Wired magazine this year that his exercise regime largely consists of squats while brushing his teeth.
“Greatness comes from character. And character isn’t formed out of smart people, it’s formed out of people who suffered,” Huang told students at Stanford University earlier this year, adding: “I use the phrase ‘pain and suffering’ inside our company with great glee.”
Huang’s commitment to pain concluded in him getting a tattoo of Nvidia’s logo when the company’s share price hit $US100 in 2014, the pain of which made him cry. He has attributed his resilience to being sent away to boarding school in Kentucky as a nine-year-old, where he was subjected to relentless racist bullying.
Most of Nvidia’s employees today would say that any suffering has been worth it. The company’s 3500 per cent stock price rise in the last five years means that even staff with relatively short careers at the company have enough paper wealth to retire. One mid-level employee who worked for the company for 18 years recently left with a $US62 million nest egg.
Huang became the world’s 12th richest person last week, with a personal wealth of $US119 billion.Credit: Bloomberg
Huang himself is not prone to outrageous displays of wealth, beyond the $US9000 leather jackets – at least by the standard of most tech billionaires. He has spent around $US55 million on property including the Hawaii retreat but is said to drive a Tesla to work despite owning a collection of Ferraris.
Some caution may be warranted. Nvidia briefly overtaking Microsoft and Apple in value has been seen as the crest of a new dotcom bubble. The company’s shares have since retreated.
Its profits, while growing strongly, are less than half of Apple’s, and are built on runaway spending on AI infrastructure that has yet to yield tangible profits for many of its customers. The company’s ascent has drawn parallels with Cisco, the networking equipment company that overtook Microsoft during the dotcom bubble but never recovered.
Loading
Most of Wall Street believes the only way is up, however. Sam Altman, the chief executive of OpenAI, has suggested building an artificial superintelligence that can outsmart humans will cost some $US7 trillion. A substantial portion of that would be likely to end up with Nvidia.
Last week some analysts predicted a valuation of $US5 trillion within a year, a level that would see Huang’s own net worth overtake Bill Gates’ and see him competing for the title of world’s richest man. Then it really would be time to party.