Part 3/12:
Meanwhile, Tesla is also developing its own supercomputing infrastructure, constructing a high-density, water-cooled supercomputer cluster in Buffalo, New York. Elon Musk has indicated that Tesla aims to reach roughly 85,000 H100 GPU equivalents by the end of 2024, with plans to massively scale up by fall 2025—potentially exceeding the size of XAI’s anticipated cluster. Although these efforts remain independent, the parallel development of large GPU clusters underscores the critical importance of supercomputing power in advancing AI.