DeepSeek Secretly Develops Inference Chip to Reduce Nvidia Reliance
量子位 · wechat · 2026-07-08
According to Reuters, DeepSeek is secretly developing an in-house AI chip designed for inference rather than training. The project started about a year ago and is currently in the early stages; they have contacted chip design firms, foundries, and memory suppliers, and are quietly hiring chip design engineers. The R1 base was trained on Nvidia H800s, later shifting to Huawei Ascend, and V4 has been adapted for Ascend. The UE8M0FP8 data format introduced in DeepSeek-V3.1 is believed to be specifically designed for the next generation of domestic chips. However, developing a competitive chip requires years and massive capital, and success is not guaranteed. Supporting this plan is an initial external funding round of about 51 billion RMB (approx. 7.4 billion USD) completed in June 2026, with a post-investment valuation of 52–59 billion USD. The funds will be used to expand computing centers dominated by domestic chips, develop proprietary chips, expand the team, and hire IDC planning engineers for data center construction in places like Ulanqab, Inner Mongolia.
More from Venture
- AI is collapsing startup and operating costs as U.S. business formation hits a record — sanjaykalra · 2026-07-21
- Soxton lands on investors’ 2026 legal-startup watchlist with $2.5M raised — ashleymayer · 2026-07-21
- A media-buying team says one AI employee scaled 5 campaigns to 60 with the same staff — aryanXmahajan · 2026-07-21
- Dimension launches an $800M third fund and says it now manages $1.65B — chaitjo · 2026-07-21
- Oracle could face a $7B collateral bill for its Wisconsin data center — 1vuio0pswjnm7 · 2026-07-21
- AI is cutting costs faster than it is creating new revenue — kevinkern · 2026-07-21