.png){kind=logo}
SiFive, a chip design startup backed by NVIDIA, has secured $400 million in a new funding round. Pushing the valuation to $3.65 billion. The funding round was led by Atreides Management. It also included major investors such as Apollo Global Management, D1 Capital Partners, Point72, T. Rowe Price, and Sutter Hill Ventures. The participation highlights rising confidence in alternative chip technologies beyond traditional systems.
Founded in 2015 by engineers from the University of California, Berkeley, SiFive is built on the RISC-V architecture. This approach differs from the chip designs commonly used by Intel and ARM.
RISC-V is open and flexible. It allows companies to customize chip designs to their needs without relying on closed systems. The technology is ideal for businesses seeking greater control and lower costs.
SiFive does not produce chips itself. Instead, the company licenses its designs to customers who build their own processors. This model is similar to how ARM operated in its early years. Recently, the latter has also moved into manufacturing, launching its own AI chip with Meta.
The new funding will help SiFive expand into AI data centers, a fast-growing segment driven by rising demand for artificial intelligence. The company is developing CPUs. These will work with NVIDIA’s ecosystem, including CUDA software and NVLink Fusion systems. The setup would allow different processors to work together in large AI server environments. Hence, improving performance and flexibility.
Also Read: Weekly Funding Roundup: Sarvam AI Nears $350M, Perplexity Launches $1M Challenge.
This move also reflects NVIDIA’s broader strategy. While rivals like AMD and Intel are competing in GPUs, the Jensen-led company is backing a new player focused on CPU design.
SiFive last raised funds in 2022 at a lower valuation. The jump to $3.65 billion shows how quickly interest in AI hardware and open chip systems is growing. As demand for scalable computing rises, companies like SiFive may play a bigger role in shaping the future of AI infrastructure.
