Fanhanshu raises $15M angel round for AI-native chip design
What's the deal? Shanghai-based startup Fanhanshu has raised nearly US$13.8M in an angel round for its AI-native chip design platform. The Shanghai Future Industry Fund led, with Shanghai Xintou, Shanghai Technology Innovation Group, Fuhai Chongchan Fund, Xuhui CapitalDealroom has a profile for this one. Try Dealroom →, Kuanqiao Hengsong, and Infinity FundDealroom has a profile for this one. Try Dealroom → joining.
What's the endgame? Fanhanshu's platform takes a chip design from a natural-language description through to fabrication. Using multiple AI agents, it handles architecture planning, RTL code generation, verification, physical layout, and tape-out, producing a GDS file ready for manufacturing.
Why now? As chip features approach physical limits, design complexity is rising faster than the supply of engineers. Founder and chief executive officer Zhou Jingyi argues that chip design is a strong fit for AI agents because the feedback is deterministic — code either compiles, simulations either pass, and timing either closes.
The money will fund training of its domain-specific model, iteration of its tool chain, hiring across chip engineering and AI, and joint delivery with customers. Fanhanshu, founded in March, has completed several rounds in seven months.
What's the traction? The company says it has helped a customer complete a commercial tape-out, with the chip returned, powered up on first try, and now in volume production. It claims design cycles can be cut to under a third of traditional timelines.
What could go wrong? Those results rest largely on the company's own account and a single production chip. Industry reporting notes that fully autonomous, end-to-end chip design remains early, with limited evidence of reliable production use.
The signal: The round reflects a global push to apply AI agents to chip design, where overseas startups such as ChipAgents have drawn funding and EDA incumbents including Synopsys, Cadence, and Siemens are building agent tools. For investors, the bet is whether domestic AI-native EDA can move from demos to dependable, scalable engineering.
Read more: finance.sina.com.cn
Image credit: Generated with Gemini