Elon Musk Says He's Building a Chip '2-3x Better Than Nvidia' at 10% the Cost. Should Nvidia Investors Be Worried?

The world's first trillionaire has the world's biggest company in his crosshairs.

Thanks primarily to his stakes in SpaceX (SPCX +12.34%) and Tesla (TSLA 1.41%), Elon Musk's net worth now tops $1 trillion. Musk has pioneered innovations in electric vehicles. He has successfully commercialized rocket launches. And now he plans to take on Nvidia (NVDA 1.49%).

Musk told Baron Capital founder Ron Baron that he's building an artificial intelligence (AI) chip that will be "two to three times better than Nvidia" and at 10% of the cost. Should Nvidia investors be worried?

Musk wasn't joking about his plans to go head-to-head against Nvidia. He told Baron in the virtual interview, "I have the entire physical design of the chip laid out in memory -- I can visualize the whole thing." And he has already revealed how he intends to build this superchip.

In March 2026, Tesla announced plans to collaborate with SpaceX and xAI (now part of SpaceX) to build TeraFab, billed as "the largest chip manufacturing facility ever." Intel (INTC 6.33%) joined the team in April.

Tesla already designs its own AI inference chips but outsources manufacturing. To achieve Musk's vision, it will need highly specialized fabrication equipment. That's where ASML (ASML 2.89%) enters the picture.

Musk spoke with ASML CEO Christophe Fouquet last week in an event exclusively for ASML employees. This wasn't their first conversation, though. Fouquet confirmed in May that he had talked with Musk about TeraFab and SpaceX, stating, "He's very serious about all those projects."

But could Musk's proposed chip really threaten Nvidia's dominance in the market? Maybe, but it won't be easy.

Developing AI chips that are two to three times better than Nvidia's industry-leading Blackwell GPUs is possible with breakthroughs in architecture, specifically in memory bandwidth. However, doing it at one-tenth the cost could prove daunting, to say the least.

Perhaps the most difficult hurdle for Musk's Tesla/SpaceX alliance is that the goalpost will continually move. Nvidia isn't resting on its laurels. The chipmaker is investing heavily in developing more powerful, cost-effective chips. As a case in point, Nvidia's new Vera Rubin chips deliver up to 10 times more tokens per megawatt than the company's GB200 chips based on the Blackwell architecture.

Another factor to consider is that TeraFab may not join Nvidia's competitors. Even if Musk successfully builds a chip superior to Nvidia's, SpaceX and Tesla could be the only customers.

Tesla's announcement about the chip manufacturing facility referenced SpaceX's enormous demand for launching "millions of tons of mass into orbit" and the massive number of chips needed for Tesla Optimus robots. The post on X (formerly Twitter) stated that the need is "more than all the chip manufacturers in the world combined can provide today, or even by 2030 (based on projected production growth)."

Should Nvidia investors be worried about Musk's ambitious strategy? I don't think so. The trillionaire has a history of sometimes overpromising and underdelivering. An idea in his mind is a long way from a real product that could dethrone Nvidia.

That said, Musk shouldn't be ignored. Despite his mixed track record, he has still managed to accomplish feats that some viewed as impossible. And if Musk makes significant progress in building a super-fast, super-cheap AI chip, the threat alone could hurt Nvidia's margins.

In some ways, though, Musk's plans to build a new chip could be viewed as good news for Nvidia. In particular, it may validate Nvidia CEO Jensen Huang's prediction that his company will generate at least $1 trillion in sales through 2027. If Musk anticipates such a massive need for SpaceX and Tesla over the next few years, Huang's forecast may not be far-fetched.