Tesla's Dojo Supercomputer Set for Launch Next Month

Posted on June 21, 2023

Tesla has announced that its highly anticipated Dojo supercomputer, touted as a game-changer for the company's autonomous driving initiatives, is scheduled to go into production next month. The announcement came via the company's new Tesla AI Twitter account.

Dojo, a custom-built supercomputer, has been designed specifically for artificial intelligence (AI) machine learning. The platform's primary focus is video training, processing the extensive video data from Tesla's substantial fleet of vehicles. This data is crucial for training the neural networks that power the company's autonomous driving technology.

Although Tesla already operates one of the world's most powerful NVIDIA GPU-based supercomputers, Dojo represents a significant step up. The new system utilizes chips and infrastructure developed in-house by Tesla, promising to enhance the company's ability to train neural networks with video data further.

During AI Day in 2021, Tesla initially unveiled the Dojo supercomputer. However, at that time, the company only had its first chip and training tiles, and was still in the process of constructing a full Dojo cabinet and cluster, also known as an "Exapod". A year later, Tesla revealed progress, including a completed system tray, and projected a fully operational cluster by Q1 2023.

Following the end of Q1 2023, Tesla confirmed via its new Tesla AI Twitter account that Dojo is due to begin production in July.

Based on the provided timeline, Tesla will continue adding trays and cabinets to the supercomputer system, with the ambition to make it one of the world's most powerful supercomputers by early 2024. Furthermore, Tesla has hinted at plans to exceed these capabilities significantly by the end of the following year.

With Dojo, Tesla is expected to finally unlock the full potential of its vast real-world driving scenario database, gathered from millions of vehicles over several years. The company anticipates that the ability to leverage this data fully will propel its autonomous driving capabilities to new heights.