Yahoo Finance

Naveen Rao on why NVIDIA needs an operator, not a CFO

Naveen Rao· Founder & CEO at Unconventional AI
·~8 min·English·Yahoo Finance
GPUOpen SourceAI InfrastructureBusiness Strategy
TL;DR

Naveen Rao argues NVIDIA's next CEO must be a deep-technology operator who can make long-horizon bets and read markets - not a finance-first leader - because he sees the CUDA-and-ecosystem advantage that built NVIDIA as what a numbers-only successor could put at risk.

01Core Mental Model

The CEO Job Is Not the CFO Job

Rao warns that running NVIDIA is a product-and-technology job, not a finance job, and an interim CFO who becomes the permanent CEO would be a very bad outcome.

the CEO job is not the CFO job. They're very different

— Naveen Rao, Yahoo Finance
Key Insight
Rao is not attacking Colette Kress personally - he concedes she has been at NVIDIA a long time. His worry is structural: a finance-shaped role rewards defending the numbers, and that instinct is the opposite of the one NVIDIA was built on.

02Proof by History

The 2014 Bet a CFO Would Never Make

When Jensen Huang committed NVIDIA to AI at GTC in 2014, the company already had a safe graphics market and no proof the AI bet would pay off - a long-horizon move Rao doubts a numbers-first leader would have made.

I don't know that a CFO would have done such a thing, right? Because it was not short-term financially driven. That was a long-term bet.

— Naveen Rao, Yahoo Finance
Key Insight
The AI pivot only looks obvious in hindsight. Rao's point is that the exact bet that made NVIDIA the most valuable chip company is the kind a spreadsheet would have rejected - so the trait you must hire for is the willingness to make it again.

03The Real Product

Not a Chip Company - an Ecosystem

NVIDIA's value is the ecosystem it funds and nurtures - CUDA, open-weight models, infrastructure - and that spend only looks smart if the leader correctly bets the demand will follow.

if you create that ecosystem and the demand comes, then it was a very smart move. If the demand never came, then it's not a smart move.

— Naveen Rao, Yahoo Finance
Key Insight
The nickname 'bank of Jensen' frames NVIDIA's ecosystem spending as generosity. Rao reframes it as a wager: it is only wise in retrospect, so the CEO's core skill is judging which ecosystems will summon the demand that pays the spend back.

04The Limit of the Analogy

Why Tim Cook Doesn't Transfer

Rao grants that Apple did well under an operator who optimized its supply chains, but argues NVIDIA is a deep-technology company whose next leap cannot be financially engineered the way operations can.

it's not something you can kind of financially engineer your way into. You have to understand the technology deeply.

— Naveen Rao, Yahoo Finance
Key Insight
The Tim Cook story is the standard reassurance when a visionary founder exits. Rao accepts it halfway: an operator can run a company whose value comes from execution, but NVIDIA's value keeps coming from technology bets, and no amount of operational polish produces the next one.

05Who Can Run It

The Bar Is Wider Than Being a Great Technologist

Rao says the successor need not be external and knows the names in play - Ian Buck, who built CUDA, and ARM's Rene Haas - but being a brilliant technologist is not the same as being able to make the company-level calls.

is he the one who's capable of making these big calls? I think that remains to be seen.

— Naveen Rao, Yahoo Finance
Key Insight
Rao declines to anoint anyone, and that restraint is the point: credentials like 'built CUDA' or 'turned ARM around' prove narrow excellence, while the job demands reading hardware, markets, AI's direction and open-source dynamics at once - a rarer combination than any single resume shows.

06The Clock

Close It in Three to Four Months

Even though the hardware roadmap is set years out, Rao calls the search urgent because NVIDIA is famous for fast roadmap changes and its talent base cares who is leading it.

I actually think it's pretty urgent and I would say like it needs to be closed probably in the next 3 or 4 months.

— Naveen Rao, Yahoo Finance
Key Insight
The intuitive read is that locked-in silicon buys NVIDIA time. Rao inverts it: the company's edge has been making sharp, real-time roadmap changes like the V100's tensor cores, and only a decisive leader can keep making those - so a drawn-out search is itself a risk to the product.

07Keeping the People

Money Doesn't Keep the Team - Mission Does

Rao argues NVIDIA's now-wealthy engineers stay for the mission, not the paycheck, so the wrong kind of CEO risks losing the very talent the company's edge depends on.

mission is super critical and that's how you get talent.

— Naveen Rao, Yahoo Finance
Key Insight
A team that has already made its money is the hardest to retain with more money. Rao points to the people who stayed through NVIDIA's lean years twelve years ago as his evidence that belief in the direction, not the paycheck, is the real glue - the kind a finance-first CEO cannot credibly supply.