Tesla announced that it has invested in a massive computer consisting of 10,000 Nvidia H100 GPUs specifically designed to support artificial intelligence workloads. With the supercomputer developed under the instructions of Elon Musk, large amounts of data received from autonomous vehicles will be processed. The system came online this week. Tesla is putting so much effort into reaching the point where its vehicles can be considered fully autonomous.
Features of Elon Musk’s new supercomputer
In July 2023, CEO Elon Musk announced that the firm will invest $1 billion to build the Dojo supercomputer next year. Based on Tesla’s own technology, Dojo includes the D1 chip equipped with 354 custom CPU cores. Each training tile module (training AI) consists of 25 D1 chips, and the basic Dojo V1 configuration contains a total of 53,100 D1 cores.
The firm also created a supercomputer equipped with 5,760 Nvidia A100 GPUs in June 2012. But the company’s latest investment in 10,000 of its H100 GPUs dwarfs the power of this supercomputer. Valued at more than $300 million, this AI cluster will deliver peak performance of 340 FP64 PFLOPS for technical computing and 39.58 INT8 ExaFLOPS for AI applications.
A hardware expert pointed out that Tesla’s power is actually more than the Lenoardo supercomputer offers, making it one of the most powerful computers on the planet. Nvidia’s GPUs currently power many of the world’s leading productive AI platforms. These GPUs installed in servers have many other uses, from medical imaging to creating weather models.
Tesla hopes to use the power of these GPUs to use the large amounts of data it has more efficiently and effectively to create a model that could revolutionize the field of artificial intelligence. While many businesses generally rely on infrastructure hosted by companies like Google or Microsoft, Tesla’s supercomputing infrastructure is entirely within the company itself.
So what do you think about this issue? You can share your thoughts with us in the comments section.
33107
