{"id":46690,"date":"2025-02-10T13:47:06","date_gmt":"2025-02-10T18:47:06","guid":{"rendered":"https:\/\/mjtsai.com\/blog\/?p=46690"},"modified":"2025-02-10T13:47:06","modified_gmt":"2025-02-10T18:47:06","slug":"deepseeks-true-training-cost","status":"publish","type":"post","link":"https:\/\/mjtsai.com\/blog\/2025\/02\/10\/deepseeks-true-training-cost\/","title":{"rendered":"DeepSeek’s True Training Cost"},"content":{"rendered":"

Anton Shilov<\/a>:<\/p>\n

SemiAnalysis<\/a> reports that the company behind DeepSeek incurred $1.6 billion in hardware costs and has a fleet of 50,000 Nvidia Hopper GPUs, a finding that undermines the idea that DeepSeek reinvented AI training and inference with dramatically lower investments than the leaders of the AI industry. <\/p>

<\/div>

DeepSeek operates an extensive computing infrastructure with approximately 50,000 Hopper GPUs, the report claims. This includes 10,000 H800s and 10,000 H100s, with additional purchases of H20 units, according to SemiAnalysis. These resources are distributed across multiple locations and serve purposes such as AI training, research, and financial modeling. The company’s total capital investment in servers is around $1.6 billion, with an estimated $944 million spent on operating costs, according to SemiAnalysis.<\/p><\/blockquote>\n\n

Yazhou Sun and Tom Mackenzie<\/a>:<\/p>\n

\n

The notion that China’s DeepSeek spent under $6 million to develop its artificial intelligence system is “exaggerated and a little bit misleading,” according Google DeepMind boss Demis Hassabis.<\/p>\n

[…]<\/p>\n

DeepSeek “seems to have only reported the cost of the final training round, which is a fraction of the total cost.”<\/p>\n<\/blockquote>\n\n

Previously:<\/p>\n