Part 8/14:
Tesla is investing heavily in computational infrastructure to support this AI revolution. Musk revealed that Tesla plans to deploy a $250 million, 10,000-unit Nvidia H100 GPU cluster, with total expected annual compute spending reaching around $2 billion. This level of investment underscores Tesla’s commitment to training massive neural networks capable of understanding complex driving scenarios.
Tesla’s efforts are likely to produce some of the largest training datasets in the world. Tim Zaman, a Tesla employee, suggested that these datasets give Tesla an unprecedented advantage in developing robust, real-world AI systems.