DeepSeek developing own AI chip

Chinese AI startup DeepSeek is developing its own AI inference chip to reduce reliance on Nvidia and Huawei, according to Reuters. The chip is designed for inference, not training, and the project is in early stages. DeepSeek has hired chip engineers and held discussions with chip-design, foundry, and memory companies. The move comes amid US export restrictions on advanced chips to China. DeepSeek also raised $7 billion in its first funding round, valuing it at $52-59 billion, with investors including Tencent, CATL, JD.com, NetEase, IDG Capital, Monolith Management, and China's National AI Industry Investment Fund. Founder Liang Wenfeng contributed $3 billion. The company plans to use funds for R&D and computing infrastructure. DeepSeek's V4 model was adapted for Huawei's Ascend chips, and orders for Huawei's Ascend 950 surged. Huawei holds about half of China's $50 billion AI chip market but faces competition from Alibaba Group, Baidu, and now DeepSeek. OpenAI unveiled its custom inference chip Jalapeno with Broadcom, and Anthropic is exploring its own chips. Nvidia shares slipped 1.6% on the news.