Rivian Doesn't Need Nvidia for Self-Driving Cars. Should Nvidia Investors Be Worried?

Rivian's (NASDAQ: RIVN) Autonomy & AI Day last week put a wrinkle into the Nvidia (NASDAQ: NVDA) growth story.

Rivian, which specializes in electric trucks and SUVs, said it plans to ship a new autonomy computer on its R2 vehicles starting at the end of 2026, built around an in-house chip designed to run its self-driving software.

Rivian is not a major Nvidia customer, so this is not about lost revenue today. It's about a pattern: More companies are looking for ways to rely less on Nvidia's pricey AI (artificial intelligence) chips -- and that trend matters more when Nvidia's stock's valuation already assumes years of continued dominance.

Image source: Rivian Automotive.

Rivian is building its own silicon

During its inaugural Autonomy & AI Day, Rivian said its updated hardware platform includes an in-house inference chip ("inference" is the step where a trained model makes real-time predictions in the vehicle).

"I couldn't be more excited for the work our teams are driving in autonomy and AI," said Rivian founder and CEO RJ Scaringe in the company's press release about its new custom silicon. "Our updated hardware platform, which includes our in-house 1600 sparse TOPS inference chip, will enable us to achieve dramatic progress in self-driving to ultimately deliver on our goal of delivering [level 4 self-driving]."

Translation: Scaringe believes the electric-vehicle company's new inference chip will help Rivian achieve driving without any human intervention.

For Nvidia investors, the key point is not Rivian's timing or feature list. Instead, it's the fact that it can meet its needs with a custom chip, rather than relying on a computing solution from Nvidia.

Nvidia will be fine without Rivian

Nvidia, of course, is still in a different weight class. Its fiscal third-quarter revenue was $57.0 billion. Further, reflecting how central its data center business has become to the AI buildout, data center revenue represented the bulk of this revenue, coming in at $51.2 billion -- up 66% year over year. This towers over Rivian's third-quarter revenue of nearly $1.6 billion.

Driving home Rivian's immaterial positioning relative to the AI chip giant, Nvidia's automotive and robotics segment remains small relative to the whole company, with third-quarter revenue of $592 million.

Still, Rivian's move could be concerning when viewed alongside similar moves from other tech companies focusing on in-house AI silicon solutions.

Alternatives are multiplying

The bigger issue is that Rivian is far from alone in trying to control its AI costs.

Alphabet (NASDAQ: GOOG) (NASDAQ: GOOGL) has built its own tensor processing units, or TPUs, and it offers them through Google Cloud as an alternative to Nvidia chips. TPUs are custom-designed AI accelerators, meaning chips built specifically to run machine learning software, and they are used for both training and inference.

Amazon (NASDAQ: AMZN) is doing something similar inside its cloud computing arm, AWS. The company has been developing its Trainium chips for training AI models, and it recently announced that its new Trn3 UltraServers can scale up to 144 Trainium3 chips for large workloads.

While there is no perfect substitute to Nvidia's AI graphics processing units (GPUs), Rivian is the latest in a growing list of companies that has come up with ways to be less dependent on Nvidia.

Sure, Nvidia can keep growing while alternatives to its chips expand, but growth could slow, and margins could compress even if competition ramps up only marginally; pricing power tends to soften once customers have credible substitutes.

Nvidia is still the leader in the space -- and demand for its latest platforms is surreal, with management emphasizing in Nvidia's fiscal third-quarter results that its Blackwell sales were "off the charts" and its "cloud GPUs are sold out."

Still, given Nvidia's price-to-earnings ratio of 44, investors have priced in continued rapid growth without margin compression for years to come. With a valuation like this, small shifts in buyer behavior could spook investors. And to be clear: Rivian building its own chip won't decide Nvidia's future, but it fits the same build-versus-buy logic playing out at the biggest cloud companies. Investors, therefore, should treat that as a real risk -- and something worth keeping an eye on.

Daniel Sparks and his clients have no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Alphabet, Amazon, and Nvidia. The Motley Fool has a disclosure policy.