Nvidia's push to dominate self-driving cars

Catch up quick: Nvidia has been involved in autos for decades, first with technology to design 3D virtual models, then by providing chips for dashboard infotainment systems, and later for driver-assist systems.

• Its digital simulation technology even helps carmakers test virtual assembly lines before a shovel hits the ground.
• Its automotive business is still tiny compared to the billions it rakes in from AI, but it's positioning itself now to become the industry's foundational platform.

Zoom in: Nvidia's Hyperion platform is a modular suite of hardware and software products that any automaker, robotaxi operator, trucking company or startup can use to build an AV system without having to build it from scratch.

• It's an open platform, meaning customers can select some or all of Nvidia's tech, whether they're developing an advanced Level 2 assisted-driving system or a fully autonomous Level 4 car or robotaxi.

The big picture: Tesla and Waymo (owned by Google parent Alphabet) have deep pockets to invest billions of dollars in proprietary AV systems.

• For other players facing global economic threats and huge technology shifts, Nvidia's tech platform offers a shortcut to autonomy.
• "How can anyone compete with Google and Tesla on their own? It's too much of an investment, and so we're trying to build a platform to make the industry, the ecosystem, successful," Ali Kani, head of Nvidia's automotive team, tells Axios.

State of play: Mercedes' new CLA luxury compact is the first vehicle to use Nvidia's full-stack AV software for its enhanced Level 2 driver-assistance feature.

• With the push of a button, attentive drivers can navigate point-to-point without touching the steering wheel, similar to Tesla's FSD assisted-driving technology.
• "What you're seeing in the case of Mercedes‑Benz is, hey, it's okay to partner right now," said Kani. "We take the investment for them."

Nvidia's plug-and-play system can also be used for fully autonomous vehicles.

• It's partnering with Uber, for example, to deploy 100,000 Nvidia-powered robotaxis from various manufacturers on the ride-hailing network starting in 2027.
• A number of companies, including Stellantis, Lucid, Volkswagen, Mercedes and Zoox, are adopting the Level 4 system, along with leading Chinese automakers BYD and Geely.

What we're watching: With the breadth of offerings, Nvidia's bet is that it can lock customers into the Nvidia AV ecosystem long term.

• "Hyperion is 'Wintel' for automotive," said Kani, referring to the dominant Microsoft-Intel platform in the 1980s and '90s.

Reality check: Analysts are skeptical that Nvidia will be able to convince the world's automakers to standardize on its AV platform.

• "Having a monopoly is rarely a good thing for any industry. It certainly was not good for the PC industry," said Sam Abuelsamid, vice president of market research for Telemetry, a Detroit-based research and strategic communications firm.
• He noted supply-chain dependency as a particular risk for auto companies.

Nvidia's chief rivals, Qualcomm and Mobileye, are going after the same market, offering alternatives to Nvidia's vertically integrated approach.

• Qualcomm, for example, offers automakers less expensive and more energy-efficient chips that link automated driving with connected-car technologies.
• Mobileye's advantage is leveraging the driver-assistance software, mapping and safety systems it has already deployed on millions of vehicles.

The bottom line: Automakers are juggling a long list of issues, not the least of which is choosing which AV architecture to deliver their self-driving future. Nvidia is betting on a big piece of the action.