We Don’t Predict the Next Wave; We Back the People Who Create It

What NVIDIA's thirty years of preparation teaches us about backing founders

People often think of NVIDIA as a company the AI wave made. I think it's the opposite: the wave found a company that spent thirty years preparing for it.

In The NVIDIA Way, Tae Kim reveals the company’s key principles – which have remained unchanged since its 1993 founding – that prepared it for (and I would even argue helped create) the AI wave. What struck me was how these principles clearly mirror the often intangible qualities we look for in the founders we back at Penny Jar. It’s a striking reminder that though technology is constantly reinvented, the hallmarks of the best founders remain eerily consistent.

Three things we look for that are brought to life in the book:

1. Talent compounds. This is always top of mind for us when considering an investment, and an easy heuristic we use is: would the smartest person we know want to work with this team? A founder’s team is the center of gravity, with an impressive (often underappreciated) pull. Every exceptional hire makes the next one easier to close. NVIDIA’s proof of concept: In 1997, Jensen convinced John Montrym, the chief engineer at Silicon Graphics, to join NVIDIA. Revered by engineers, Montrym was a watershed hire: every time NVIDIA posted a JD, resumes and interview requests poured in from Silicon Graphics (and others). Everyone wanted to work with the best people on the best technology, and NVIDIA had both. ‍
2. Speed is a muscle. At Penny Jar, we call this “moving with pace” or having a “fast clock speed.” Our heuristic is: how much has changed since the last time we spoke? With the best founders, each check-in feels like a leap forward, even if it’s only been a few days. NVIDIA’s proof of concept: The chip industry shipped on a two-year cycle, and NVIDIA decided to ship every six months. Instead of the standard loop of designing, printing, testing, and iterating a chip, they built a front-end emulator to debug as much as possible in software before committing to hardware. Jensen pushes everyone to work at the “Speed of Light," constrained only by the laws of physics, not internal politics or financial caution. Once built, the muscle never went away, and speed forgives mistakes: a flop is a fix six months out, not two years.‍
3. Make bold bets. We look for this in all of our founders, people who have insights, not just ideas, and are willing to place their bets accordingly. A simple heuristic is: when I spend time with a founder, do they say things that make me want to pull out a notebook and write them down? When you meet hundreds of founders each year, that urge is indicative of someone who is sharing something truly unique, an insight you can bet on. NVIDIA's proof of concept: whether it was gaming, scientific computing, or AI, NVIDIA followed the same pattern: watch closely who's using your product and how, spot the seeds barely sprouting, and throw the full weight of the company at them before it’s obvious to the rest of the market. Case in point: NVIDIA bet on AI years ago, despite Wall Street’s skepticism, because of the insights it gleaned.

In moments like these, when it feels like a new model drops every few hours, it's tempting to think that keeping up means abandoning the fundamentals in favor of a set of shiny new operating principles. It's the opposite. The companies that thrive in every technology shift are the ones that get the timeless things right: the best people, a relentlessly fast clock speed, and the conviction to make bold bets on their insights. That's what we look for at Penny Jar.

We don't try to predict the next wave; we back the people who create it.