NVIDIA's DRIVE AGX Thor Powers WeRide's HPC 3.0 Platform, Reducing Autonomous Vehicle Costs by 50% and Paving the Way for Large-Scale Commercial Deployment.

WeRide has launched the HPC 3.0 high-performance computing platform, jointly developed with Lenovo and powered by NVIDIA's DRIVE AGX Thor chips. The platform reduces autonomous driving suite cost by 50% and is the most powerful computing platform available to support L4 autonomy. It features dual NVIDIA DRIVE AGX Thor chips and is certified to AEC-Q100, ISO 26262, and IATF 16949 standards. HPC 3.0 is built for 10 years or 300,000 km of use and is suitable for deployment across international markets.

WeRide has made a significant stride in the autonomous driving sector by unveiling the HPC 3.0 high-performance computing platform, developed in collaboration with Lenovo and powered by NVIDIA's DRIVE AGX Thor chips [1]. The platform, which debuts in WeRide's latest-generation Robotaxi GXR, marks the world's first mass-produced Level 4 (L4) autonomous vehicle built on NVIDIA DRIVE AGX Thor.

HPC 3.0 delivers up to 2,000 TOPS of AI compute, making it the most powerful computing platform available for L4 autonomy. The platform features dual NVIDIA DRIVE AGX Thor chips and is certified to AEC-Q100, ISO 26262, and IATF 16949 standards. This robust certification ensures exceptional reliability, with a failure rate below 50 FIT and a mean time between failures (MTBF) of 120,000 to 180,000 hours. Additionally, HPC 3.0 is built for 10 years or 300,000 km of use and is capable of operating in extreme temperatures from -40°C to 85°C, making it suitable for deployment across international markets [1].

The platform's cost efficiency is notable, with a 50% reduction in autonomous driving suite costs and an 84% lower total cost of ownership (TCO) compared to its predecessor, HPC 2.0. This significant reduction in costs is achieved through core upgrades and system integration, which cut mass production costs to a quarter of HPC 2.0. Key modules such as Ethernet gateway, CAN gateway, inertial navigation, and collision detection are consolidated, reducing both production and maintenance expenses [1].

The collaboration between WeRide, Lenovo, and NVIDIA represents a substantial technological leap in autonomous driving. The HPC 3.0 platform's dual NVIDIA DRIVE AGX Thor configuration, running the safety-certified DriveOS, is built on Lenovo's AD1 L4 autonomous driving domain controller. This partnership has been instrumental in accelerating the innovation and commercialization of autonomous driving solutions globally [1].

With the successful deployment of NVIDIA DRIVE AGX Thor on WeRide's Robotaxi GXR, the company is positioned to accelerate its international expansion beyond its current 2,000+ days of safe Robotaxi operations. The cost efficiencies and performance improvements should significantly compress the timeline for profitable autonomous mobility services at scale [1].

References:
[1] https://www.stocktitan.net/news/WRD/we-ride-teams-up-with-lenovo-to-launch-100-automotive-grade-hpc-3-0-0hqw7efsv0c2.html