China's Didi recently announced cooperation with Nvidia and the two parties will jointly develop autonomous driving and cloud computing solutions.

With the cooperation, Didi will use Nvidia GPU to train machine learning algorithm in its data center and it will adopt Nvidia Drive to provide reasoning ability to its L4 self-driving car. In August 2019, Didi announced to upgrade its autonomous driving division into an independent company and would implement extensive cooperation with industrial chain partners.

As a part of Didi's autonomous driving AI processing, Nvidia Drive will apply multiple deep neural networks to integrate data from various sensors, including camera, lidar, and radar, to realize comprehensive understanding of environment around car, so as to plan safe driving path. To train the deep neural networks, Didi will reply on Nvidia GPU data center server.

In regards to cloud computing, Didi will build AI infrastructure and launch computing, rendering, and gaming vGPU cloud servers. Didi Cloud will adopt a new vGPU licensing model.

So far, Didi Cloud has cooperated with industrial partners like Nvidia in several sectors, including transportation, AI, graphics rendering, electronic game, and education and training.