To The Moon

hey everyone, Wonton Investor here. i’m sure by now yall have seen the news that Jensen Huang has resigned as CEO of $NVIDIA Corporation(NVDA)$, effective immediately, after the affair scandal that’s been building all week. CFO Colette Kress has been named interim CEO and will keep her CFO job at the same time, Harvey Jones becomes the first ever chair of the board, and the board says it’ll look at both internal and external candidates for the permanent role. the stock dropped around 10% on the news, which hurts a bit more when you remember it was already down over 20% in the past month 🥴

so the question everyone’s asking is who actually takes over, and my (very unofficial) bet is Ian Buck. if you’ve never heard of him, don’t worry, most people outside the industry haven’t, which is kinda funny because he’s arguably the reason Nvidia is where it is today.

who is Ian Buck?

Buck did his undergrad in computer science at Princeton and his PhD at Stanford, and the story goes that around 2000 he was wiring a bunch of GeForce cards together to run Quake 3 at an absurd resolution. that experiment turned into Brook, one of the first programming languages that let people use a GPU for general computing instead of just graphics. Nvidia hired him in 2004, and he went on to start the team that built CUDA, which launched in 2007 (it was a team effort, but he’s widely called its architect).

if CUDA sounds like jargon, think of it like this: Nvidia’s chips are the trains, but CUDA is the railway track that basically every AI lab and developer in the world has already built their stations around. a competitor can build a faster train, but if it doesn’t run on the existing tracks, switching is a massive headache, and that headache is a huge part of Nvidia’s moat.

today he’s VP of hyperscale and HPC, which means he oversees the AI hardware and software sold to Microsoft, Google, Amazon, Meta and OpenAI, aka the part of Nvidia that makes most of the money. he’s also advised the White House and testified to Congress on AI, so he’s a lot less of an unknown than his public profile suggests.

why him and not Kress?

Kress is the obvious steady hand (she’s been CFO since 2013 and sits next to Jensen on every earnings call), and i think she’s a great pick for the interim. but Nvidia’s edge has always come from having a deeply technical person at the top who can see where computing is headed, and Buck is the closest thing they have to that internally. the small prediction markets on this also had Buck first and Kress second, though i wouldn’t read too much into that since the volume is tiny. and if you know Nvidia's business, CUDA is the main reason AMD and Intel struggle to keep up with Nvidia's chips. their real moat is and has always been in CUDA. AMD's software (ROCm) only started receving more investments in recent years, while CUDA has been in the works since the late 2000s, so CUDA is much more developed and well supported than other softwares on the market. does this mean that this moat can stay forever? no, but it definitely gives Nvidia an edge in the near future, and i'm willing to bet my life that Nvidia will defend CUDA with everything they have.

what could prove me wrong

he’s famously low-key and nothing like the showman Jensen is, the board might decide it wants an outsider (Arm’s Rene Haas, who used to work at Nvidia, keeps getting mentioned), and a co-CEO setup with Kress on the business side and Buck on product wouldn’t surprise me either.

this is a guess, not a recommendation, and i’m not telling anyone to buy or sell NVDA off the back of it. earnings are next week, and i care a lot more about how the company sounds without Jensen than about the numbers themselves. [disclosure: i’m long NVDA through a long-dated position]

i wrote a slightly longer note on this on my Substack if you want the full background.

peace out,

The Wonton Investor