Pony.ai to equip its robotaxi fleet with the NVIDIA DRIVE AGX Pegasus platform

To meet the massive computing demands required to bring robotaxis to market, Pony.ai will equip its robotaxi fleet with the NVIDIA DRIVE AGX Pegasus platform, the company recently announced. 

According to Pony.ai, DRIVE AGX Pegasus achieves an unprecedented 320 TOPS (trillions of operations per second) of deep learning. The platform integrates two NVIDIA Xavier SoCs and two NVIDIA Turing Tensor Core GPUs; both of which achieved top results in MLPerf Inference, the first industry-standard, independent AI benchmarks.

“As the only open, end-to-end and software-defined AV platform currently available for the robotaxi market, the NVIDIA DRIVE AGX Pegasus is an excellent solution for helping us realize our vision of bringing safe, autonomous mobility solutions to all,” says Tiancheng Lou, co-founder and CTO, Pony.ai.

Since late 2018, Pony.ai has been piloting its robotaxis—known as PonyPilot—in several geographies across the world including California and China. Pony.ai is the first company to offer a robotaxi service to the general public in California, and has been operating in Irvine, California since Nov. 2019.

“Pony.ai is helping advance the future of urban mobility, making transportation safer, more enjoyable, and more accessible to all,” says Rishi Dhall, VP, Autonomous Vehicles, NVIDIA.

“Pony.ai’s selection of the NVIDIA DRIVE AGX Pegasus platform gives them the scalable computational horsepower and sophisticated AI essential for safe self-driving systems.”