Leading high‑performance RISC‑V processor IP company SiFive has announced it will become the first RISC‑V chip designer to integrate NVIDIA’s NVLink technology, enabling high‑speed chip‑to‑chip interconnectivity and opening new paths for the convergence of open architectures with advanced AI computing.
As a chip IP provider positioned alongside Arm (owned by SoftBank), SiFive supplies core component blueprints that allow clients to create their own complete chip designs. RISC‑V serves as an open‑standard alternative to Arm’s closed architecture, attracting strong interest due to its flexibility and customization advantages. With Arm increasingly moving into chip design, technology giants such as Google and Meta have heightened their focus on RISC‑V. This integration further strengthens the competitiveness of the open architecture ecosystem.
The core value of NVIDIA’s NVLink technology lies in creating high‑bandwidth, low‑latency coherent interconnects between CPUs and AI accelerators—a critical need for AI data centers where massive AI tasks require the connection of thousands of chips working in unison to process large‑scale data seamlessly and overcome computational bottlenecks. This collaboration deeply aligns SiFive’s RISC‑V architecture with NVIDIA’s acceleration ecosystem, addressing the interconnect challenge between open architectures and leading AI chips.
SiFive CEO Patrick Little indicated that chip designs incorporating the technology are expected to reach the market in 2027 or later. Clients will then be able to connect RISC‑V CPUs with NVIDIA’s leading AI chips at high speeds, achieving performance comparable to interconnects between Intel and Arm‑based CPUs. “This is a multi‑generation commitment from both companies, and we are focused on developing and delivering the best NVLink solutions moving forward,” he stated.
This partnership marks a significant step for RISC‑V as it moves beyond microcontroller applications into high‑end AI computing. By breaking the monopoly of traditional architectures, it offers data centers heterogeneous computing solutions that combine customization freedom with high energy efficiency, reshaping the competitive landscape of AI infrastructure.
ICgoodFind : The collaboration bridges open architecture with leading‑edge computing, accelerating RISC‑V adoption in the AI sector and driving the industry toward a more diverse architectural ecosystem.
