DeepSeek Develops Its Own AI Chip

Chinese AI startup DeepSeek is developing its own artificial intelligence chip, according to three people familiar with the matter — a strategic shift that could reduce its reliance on Nvidia and Huawei semiconductors.

The custom chip is designed specifically for inference — the stage where a trained AI model generates responses for users — rather than for training new models, the sources told Reuters.

If successful, DeepSeek's expansion into semiconductor development would mark a major strategic pivot for a company widely hailed as China's AI champion. The Hangzhou-based startup rose to global fame over a year ago after releasing highly efficient AI models that surprised Silicon Valley and Washington.

Following Global Trends

DeepSeek joins a growing list of AI developers seeking greater control over their hardware. OpenAI recently unveiled Jalapeño, its first custom inference chip developed with Broadcom, while Anthropic has been weighing building its own AI chips.

Early Stage, Quiet Hiring

The effort remains at an early stage. DeepSeek has reached out to external partners and held discussions with chip-design firms, foundries, and memory companies. The project began about a year ago, one source said.

In recent months, DeepSeek has quietly increased hiring of chip-design engineers — but recruitment has been done privately, without job postings on public platforms, two sources noted.

Market Impact

Shares of Nvidia slipped about 2% in premarket trading following the news. While Huawei's offerings still lag Nvidia's most advanced chips significantly, a U.S. export ban has helped Huawei capture roughly half of China's $50 billion domestic AI chip market — a position now being challenged as rivals Alibaba, Baidu, and now DeepSeek develop their own silicon.

"DeepSeek's move underscores a broader trend: the world's leading AI labs no longer want to be at the mercy of chip suppliers," said a semiconductor analyst. "Custom silicon is becoming a competitive necessity."

Sources