Nvidia-backed SiFive hits $3.65 billion price tag for open-source AI chips

SiFive, a company founded in 2015 by UC Berkeley engineers who developed an open-source chip design, has arrived. $400 million in funding a round that makes the company worth $ 3.65 billion.

This agreement is interesting for several reasons. For one, SiFive’s RISC-V open chip design is based on a RISC processor, not Intel’s x86 or ARM, the two main types of CPUs that are powering Nvidia’s GPU computer AI empire.

Also, Nvidia was an investor in this round, along with a long list of VCs, private equity, and hedge funds. The round was led by Atreides Management, founded by former Fidelity chief executive Gavin Baker. (Atreides was also a merchant Cerebras Systems $1 billion round). Other investors in the round are Apollo Global Management, D1 Capital Partners, Point72 Turion, T. Rowe Price Sutter Hill Ventures, and others.

SiFive’s business is like Arm was in years past – it licenses them to those who modify them to meet their needs and does not sell the chips themselves. (In March, Arm changed its brand when it launched its first chip, AI chip, developed by Meta and customers including OpenAI, Cerebras, and Cloudflare.)

SiFive stands in the space with chips that are open, non-proprietary, and neutral, independent of specific customers. In fact, SiFive hasn’t raised since March 2022, Pitchbook estimates, when it brought in $175 million led by Coatue Management at a cash value of $2.33 billion. Intel Capital, Qualcomm Ventures, Aramco Ventures, were part of that round.

RISC-V has been, until recently, best known as a chip for small applications, such as embedded systems. But with this money and Nvidia’s interest, SiFive is moving towards AI data center CPUs. SiFive design it will work with Nvidia CUDA software is its own NVLink Fusionrack system that allows different CPUs to connect in Nvidia’s “AI factory”.

In other words, as rivals Intel and AMD look to compete with Nvidia’s GPU, Nvidia is backing an 11-year-old startup that can build CPUs on open and proprietary technology.