Nvidia is preparing to ship Drive Thor, its next-generation automotive chip, which the company says will be able to integrate a wide range of in-car technologies from automated driving features and driver monitoring systems to streaming Netflix in the back for the children.
Thor, which goes into production in 2025, is notable not only because it’s a step up from Nvidia’s Drive Orin chip. It also takes the place of Drive Atlan in the lineup.
Nvidia is removing the Drive Atlan system on a chip ahead of schedule for Thor, founder and CEO Jensen Huang said Tuesday at the company’s GTC event. Always in a race to develop bigger and badder chips, Nvidia is choosing Thor, which at 2,000 teraflops of performance will provide twice the computation and throughput, according to the company.
“If we look at a car today, the advanced driver assistance systems, parking, driver monitoring, camera mirrors, digital instrument cluster and infotainment system are all different computers distributed throughout the vehicle,” said Nvidia’s vice president of automotive Danny Shapiro at the press briefing Monday. “In 2025, these functions will no longer be separate computers. Rather, Drive Thor will allow manufacturers to effectively consolidate these functions into one system, reducing overall system costs.”
One chip to rule them all. A chip to help automakers create software-defined autonomous vehicles. One chip for continuous over-the-air upgrade.
Nvidia already has several automotive customers building software-defined fleets using Drive chips. For example, Volvo announced in January at the annual CES technology conference that its new automated driving features will be powered by Drive Orin. The automaker also said it will power its infotainment system with Qualcomm’s Snapdragon chip. It’s this sharing of space with competitors that probably drove Nvidia to create a more robust chip.
Zeekr is the first to raise his hand for Thor. The Chinese luxury EV startup owned by Geely has said it will use the advanced chip in its next-generation vehicles starting in 2025, according to Shapiro.
Xpeng is already using Drive Orin, the latest generation chip, for its G9 SUV, which will be able to support many advanced driver assistance functions “such as parking and driving on main and secondary streets, highways and private roads, and safe driving, entering and exiting freeways, city interchanges and toll collection routes,” Shapiro said. No doubt Xpeng who recently released their ADAS urban navigated pilot in his P5 sedan and plans to release it in the G9, will sign for the upgraded chip.
Shapiro also noted that autonomous solutions provider QCraft will begin robotics operations in China powered by Orin.
Other automakers that have previously announced the use of Nvidia’s Drive Orin include Baidu’s EV company JiDU Auto, NIO, Li Auto, R Auto, IM Motors and Polestar. It should be noted that a significant number of Nvidia’s automotive customers are based in China. While the chip maker is based in California, its chips are made, along with almost all others, in Taiwan.
Can Nvidia deliver Thor to Chinese customers?
Earlier this month, the US government imposed restrictions on exports of advanced AI chips in China, including Hong Kong and Russia. Nvidia doesn’t sell to Russia, but sanctions against China could costing the company as much as $400 million in potential sales in the third quarter. The US said the move would address the risk of the chips being used or diverted to “military end use” or “military end user” in China and Russia, but it is also a move by the Biden administration to prevent China from become a more dominant player in the core and lucrative chip manufacturing industry.
The government has restricted access specifically to Nvidia’s A100 and H100 GPUs. Fortunately for Nvidia, the company can continue to manufacture the H100 in China, although purchases from Chinese customers will be limited.
Shapiro said automotive customers would not be affected by the restrictions imposed on Nvidia’s high-end data center products, and that the company was working with Chinese customers and the US government to “present various alternatives that are not subject to same license requirements.”
Drive SIM and digital twin technology updates
At the GTC event, Nvidia also announced that its end-to-end simulation platform, I drive SIMis getting a new suite of artificial intelligence tools it calls a “neural reconstruction engine” to help test and develop self-driving vehicles.
“Using a neural engine composed of multiple [deep neural networks]”real-world rides can be recreated and reproduced accurately in simulation with the ability to change sensor configurations and locations, create new scenarios, and modify or add new aspects of other road users’ behavior,” Shapiro said.
The way it works is that AI can deconstruct a 3D scene from recorded sensor data, and that scene can then be augmented in Drive SIM with human-created or AI-generated content.
Nvidia said these upgrades will also allow automotive designers, software engineers and electronics engineers to collaborate on Drive SIM to simulate the software in the car.
The company also announced the second generation of Nvidia OVX, which will deliver immersive digital twins of cities that can be driven on Nvidia Omniverse.
BMW Group and Jaguar Land Rover are among the first clients to use OVX, Shapiro said, noting that marketing group WPP is using the Omniverse cloud to create a suite of services for automotive clients, such as personalized programmatic ads that include “perfect photo real content through virtual sets, which will save car companies orders of magnitude of costs” for expensive photo and video shoots, which are usually taken in scenic locations around the world.