The Nvidia Way

than yourself.” However, as Nvidia grew and began adding new employees at a pace of more than one hundred a month, executives understood that they would occasionally make the wrong decision. Better to correct those mistakes as soon as possible than to let them fester and harm Nvidia’s culture.

“But the reason they could do the famous papers was because they were getting the time on supercomputers,” Walker said. “Other people who had great ideas never got the time and couldn’t make their impact. It wasn’t how good your science was, but it was whether you could get access to computing time.” It was a catch-22: the only way to get supercomp

“We always have a PIC for every project. Whenever Jensen talks about any project or any deliverables, he always wants the name. Nobody can hide behind, ‘such and such a team is working on that,’ ” former finance executive Simona Jankowski said.11 “Everything has to have a name attached to it because

All morning, management provided details on its plans to expand high-performance GPU computing to new markets such as industrial and medical research applications. The company estimated that the GPU computing market would rise to more than $6 billion in a few years, even though it was near-zero at the time. In particular,

“Oh my God, we got here and we thought there was going to be a secret sauce,” one engineer said.3 “It turns out it’s just really hard work and intense execution on schedules.” It was Nvidia’s culture, in other words, that made

mine for Nvidia. In mid-2024, J.P. Morgan published survey results from 166 chief information officers, who are responsible for $123 billion in annual enterprise tech spending. The report revealed that CIOs plan to increase their AI compute hardware spending by more than 40 percent annually over the next three years, going from 5 percent of total I

former senior executive at a large software company said that he was always struck by how you could talk to multiple Nvidia employees and

As we saw in the Introduction, perhaps his favorite tool for showing the rest of the company his thought process is the whiteboard.

Even in 1993, he had to convince himself there was a $50 million annual revenue opportunity in PC graphics if he was going to leave behind steady employment and cofound Nvidia. To survive after the failure of the NV1 and NV2, he had to recalibrate Nvidia’s strategy to go after the very top of the market. Here again, the opportunity was clear: altho