Nvidia first dominated the graphics card market for video games, then the processors for artificial intelligence; now it wants to conquer the autonomous driving technology sector as well. The goal of the American company is to transform its platform for autonomous vehicles into an industry standard, much like Microsoft and Intel did with computers in the Eighties and Nineties with the “Wintel” model: Windows software on one side and Intel processors on the other.
SPREADING ARTIFICIAL INTELLIGENCE IN THE AUTOMOTIVE INDUSTRY
Nvidia is interested in spreading the use of artificial intelligence across various sectors of the “real” economy to secure new sources of revenue: it is focusing on telecommunication networks, for example, but believes the next big market for artificial intelligence is the automotive one. The company believes that many car manufacturers do not have the financial resources and technological capabilities necessary to develop advanced autonomous driving systems on their own, especially in a phase of strong margin pressure and transformation of the mobility industry.
NVIDIA IS ALREADY PRESENT IN THE AUTOMOTIVE SECTOR
In fact, as Axios reminds us, Nvidia has been present in the automotive sector for decades: initially providing technologies for three-dimensional vehicle design, then microchips for infotainment platforms, and finally advanced driver assistance systems. Nvidia also offers digital simulation tools that allow car manufacturers to “test” production lines even before the construction of factories.
The contribution to revenues from the automotive segment is still small, especially compared to the huge revenues of the data center division, but the company has big plans for the future.
WHAT IS HYPERION
The platform Nvidia is betting on to build a dominant position in the automotive market is called Hyperion. It is a modular suite composed of hardware and software that allows car manufacturers, robotaxi companies, or trucking companies to develop autonomous driving systems without having to start from scratch. The idea – essentially – is to offer a plug-and-play platform, that is ready to use but customizable: customers can decide whether to adopt only some Nvidia components or use the entire ecosystem.
Hyperion is designed both for Level 2 assisted driving systems, where the driver must remain alert and ready to intervene, and for Level 4 systems, where vehicles are fully autonomous (under certain conditions).
PARTNERSHIPS WITH MERCEDES-BENZ AND UBER
The first car manufacturer to use Nvidia’s complete stack for assisted driving will be the German Mercedes-Benz on the new Cla model, equipped with Level 2 functionalities. The agreement has great value for Nvidia, also symbolic, because it shows that car manufacturers are willing to delegate strategic technologies to an external supplier.
Nvidia is also collaborating with Uber to build one hundred thousand fully autonomous robotaxis by 2027.
THE DOUBTS
Even though Nvidia can leverage the fact that it is already the international reference for advanced processors, not everyone believes it will become the standard for autonomous driving as well.
While it is true that car manufacturers might not afford to spend billions of dollars on proprietary technologies and might therefore find it convenient to rely on ready-made solutions, some analysts think it would be risky to depend on a single technology supplier, given the general context of “fragility” in supply chains.
Competition to Nvidia, however, exists and is significant: Qualcomm offers less expensive microchips, while Mobileye (an Intel subsidiary) can leverage the fact that its software and maps are already installed on millions of vehicles. Tesla and Waymo (Alphabet group), moreover, are investing multi-billion sums in the development of proprietary autonomous driving systems.
