Hangzhou’s Nano-core Chip, which designs accelerators for AI inference, disclosed on October 28 that it had closed more than RMB 1 billion in Series C2 capital. JianTou Investment took the lead position, backed by Hangzhou Capital and Kaitai Capital among others.
The company intends to spend the money on developing products and getting ready to manufacture at scale. Coming only weeks after a Series C1 of the same headline amount in September, the pace of the raise reads more as evidence of what investors want than of any shortfall in the company’s coffers.
Nano-core launched in 2021 out of a Peking University research institute based in Zhejiang. Its chips marry three approaches: computing close to memory in three dimensions, computing inside memory itself, and RISC-V cores for control. The aim is to reduce how much weight data has to travel between memory and arithmetic, the movement that consumes most of an inference chip’s power.
Infrastructure and edge hardware both appear on the roadmap, so the design is not locked to a single socket. According to the company, all three of its first designs have finished tape-out, and sample units will go to customers ahead of initial commercial shipments early in 2027.
That schedule is the part to watch closely. Tape-out has become routine for Chinese fabless houses; what has not is shipping in volume into a customer’s finished product, which turns on foundry capacity and on software mature enough for engineers who have never targeted the instruction set before.
Choosing RISC-V is the strategic wager. It severs the design from anyone’s licensing terms while leaving the company to construct most of its own toolchain. Two checks of this size buy the time to finish that work.
