SiFive, a chip design startup based in the US, has just successfully raised 400 million USD in a new funding round, raising the company's valuation to 3.65 billion USD.
Notably, the deal involves Nvidia, showing increasing interest in open source chip technology in the artificial intelligence (AI) race.
Founded in 2015 by engineers from UC Berkeley University, SiFive emerged thanks to the development of RISC-V architecture. This is an open source chip design platform.
Unlike popular architectures such as x86 or ARM that are controlled by large corporations, RISC-V allows companies to freely customize designs, reduce costs and increase flexibility.
This capital call round is led by Atreides Management (an investment fund based in the US), with the participation of a series of large investment funds such as Apollo Global Management, D1 Capital Partners or T. Rowe Price.
Previously, in 2022, SiFive raised 175 million USD with a valuation of about 2.33 billion USD, showing a significant growth rate after only a few years.
SiFive's business model is similar to Arm in the early stages, when the company did not directly produce chips but licensed designs for customers.
Partners will customize the architecture according to their own needs, from consumer devices to data centers. However, the difference of SiFive lies in the openness and neutrality of RISC-V, not bound by the exclusive ecosystem.
For many years, RISC-V was mainly used for embedded systems or small-scale devices.
But with new capital and support from Nvidia, SiFive is strongly shifting to the field of AI data centers, where the demand for performance and customization is increasing.
According to the plan, SiFive's CPU designs will be compatible with Nvidia's CUDA software platform and NVLink connection system.
This allows RISC-V-based processors to be integrated into AI factories, where large-scale server systems specialize in processing data and training artificial intelligence models.
Nvidia's investment move is also considered a noteworthy strategy. Instead of just directly competing with competitors such as Intel or AMD in the GPU segment, Nvidia is expanding the ecosystem by supporting a completely different CPU platform. This could help the company consolidate its position in the global AI value chain.
If successful, SiFive can become a formidable counterweight to traditional chip architectures. The development of RISC-V not only opens up new choices for the semiconductor industry, but also contributes to promoting the "open" trend in technology, where businesses can flexibly build systems according to their own needs.
