Chip company SiFive secures $400 million in financing, valuing it at $3.65 billion; NVIDIA participates, betting on RISC-V

Chip company SiFive, founded by engineers from the University of California, Berkeley, recently completed a round of oversubscribed financing amounting to $400 million, raising its valuation to $3.65 billion. The funding round was led by Atreides Management, with NVIDIA participating, along with several top-tier venture capital and institutional investors, indicating high interest from the capital market in the potential of the RISC-V architecture in the AI era.

SiFive's core technology is based on the open-source RISC-V instruction set, offering a differentiated path compared to mainstream x86 and Arm architectures. Its business model follows an early strategy of Arm, which involves licensing chip designs for customers to customize, rather than directly selling chip products. However, the industry competition landscape is changing as Arm launched its own AI chip in March this year, collaborating with clients such as Meta.

In terms of funding, SiFive's previous financing dates back to 2022, led by Coatue Management with a $175 million investment, with a pre-money valuation of $2.33 billion. This round not only significantly increased its valuation but also provided ample resources for expanding into the high-performance computing field. Previously, RISC-V chips were mainly used in lightweight scenarios such as embedded systems, but SiFive is accelerating its entry into the AI data center CPU market with capital and ecosystem support.

In terms of technical collaboration, SiFive plans to make its CPU designs compatible with NVIDIA's CUDA software ecosystem and NVLink Fusion architecture, enabling integration into NVIDIA's "AI factory" system. This means that while Intel and AMD are trying to counter NVIDIA at the GPU level, NVIDIA is expanding its control over the AI infrastructure layer by investing in open-architecture CPU companies.

SiFive's funding highlights the strategic value of RISC-V in the AI computing system and reflects the trend of the chip industry moving from closed architectures to open ecosystems. As more capital and major players enter the market, the next-generation AI computing landscape centered around CPU and GPU collaboration is accelerating its transformation.