a A funny thing happened on the way to the future. It recently took place in a huge sports arena in San Jose, California, and was dubbed by some as “AI Woodstock.” But while that original music festival had attendees who were mostly high on conventional narcotics, the approximately 11,000 people in San Jose were high on the Kool-Aid so liberally provided by the tech industry.
They were gathered to listen to A keynote address at a technology conference given by Jensen Huang, the founder of computer chip maker Nvidia, who is now the Taylor Swift of Silicon Valley. Dressed in his usual leather jacket and white-soled trainers, he delivered a bravura 50-minute performance reminiscent of Steve Jobs in his heyday, albeit with slightly less slick execution. The audience also remembered the fanboys who lined up for hours to gain access to Jobs’ reality warping field, only Huang’s fans weren’t as attentive to the cues he gave them to applaud.
Still, it made for interesting viewing. Huang is a compelling speaker and he has built a remarkable company in the years since 1993, when he first sketched out his idea for Nvidia in a Silicon Valley restaurant. And the audience was in awe of him because they saw him as a man who saw the future long before they did, and hoped to catch a glimpse of what might happen next.
And in this they were not disappointed. What comes next is Nvidia’s Blackwell B200 chip, complete with its 208 billion transistors, and the family of monster machines it will enable, including a formidable rack-fitting supercomputer with nearly two miles of copper cables neatly woven together . Hear wild applause.
Watching this spectacle, the following thought occurred to me: how did a small company specializing in graphics cards for gamers become the third most valuable company in the world? And how did it end up happening so quickly? Nvidia was, after all only worth it $278 billion in October 2022 and is now worth $2.3 trilliononly behind Apple and Microsoft.
It’s a good story and no doubt someone is already working on the screenplay. But even a cursory account paints a picture of a company that, from the start, was good at anticipating the needs of a particularly demanding class of users – gamers – and eventually realized that by developing processors that could meet their needs had spawned a new kind of technology. computer: a graphics processing unit (GPU) that could perform many calculations in parallel, unlike conventional CPUs that did everything serially.
A crucial moment came in 2013, when Huang concluded that GPUs could be useful for an emerging technology called machine learning, and that the company would focus on that going forward. That was a bold gamble at the time, and initially Wall Street thought it was foolish. But when machine learning really started to take off and urgently needed parallel processing engines to perform the associated heavy computations, Nvidia hit the jackpot. If you wanted to do this kind of AI, you needed Nvidia GPUs – lots of them. And more importantly, you needed a way to make them work together seamlessly: some kind of operating system. A Stanford software genius named Ian Buck created one for Huang. They called it CUDA (for compute unified device architecture), and from then on, buying an Nvidia kit became a no-brainer for anyone looking to get into the AI ​​business.
And so Huang was virtually the only trader with a supply of ready-made shovels during the greatest gold rush in technology history. In February his company reports record quarterly revenue of $22.1 billion, up 22% from the previous quarter and up 265% from a year ago.
The Blackwell B200 chip is its latest super scoop. And in his keynote, Huang revealed what can be done with it. He rolled out the DGX GB200 NVL72 (Nvidia doesn’t do people-friendly labeling), which a powerful supercomputer with 72 Blackwell processors in one water-cooled rack. I couldn’t find any information about prices, but I guess if you have to ask, you can’t afford it.
Google can do that. This also applies to Microsoft, Meta, Oracle, Tesla, Amazon and Dell. Of what say their bosses, they are already lining up for supplies. In this case, Huang has indeed seen the future. And it works for him. Whether it works for the rest of us remains to be seen.
What I’ve read
Do not forget me
A thoughtful essay by René Walter in his Substack magazine, Good Internet, about the intrinsic hostility to cultural memory in a digital world.
The politics of technology
a great review by Bill Janeway on the online opinion site Project Syndicate from the great book by Daron Acemoglu and Simon Johnson, Power and progress.
Thomas has doubts
Jia Tolentino in the New Yorker (paywall) doesn’t like the Fat Controller in the railway stories of Thomas and friends. Me neither.