SiFive, a leader in RISC-V processor IP, will adopt NVIDIA NVLink Fusion. This integration is for its high-performance, data center solutions. This move greatly increases design options for creating integrated AI systems on the RISC-V open instruction set. AI is changing data center designs. System designers now focus on open, customizable compute platforms. These platforms can scale easily with new acceleration technologies.
AI-driven computing has reached a new stage. Now, architectural flexibility, efficient data movement, and power optimization matter as much as raw performance. The fast rise of training and inference workloads is exceeding power limits. Data center operators are rethinking how they connect and manage CPUs, GPUs, and special accelerators due to this challenge. Performance per watt and low-latency, high-bandwidth interconnects are key design needs.
SiFive’s power-efficient compute solutions adapt to changing needs. SiFive combines wide, out-of-order RISC-V cores with scalable coherent fabrics and advanced memory hierarchies. This helps cloud providers and system vendors customize CPU designs for their specific workloads. Customers gain from RISC-V’s long-term software portability. This feature prevents vendor lock-in and promotes innovation.
Also Read: Google Unveils Personal Intelligence Feature for Gemini App, to Deliver More Personalized AI Assistance
NVIDIA NVLink Fusion lets SiFive-based platforms connect directly to NVIDIA GPUs and accelerators. This connection is coherent and offers high bandwidth. This integration cuts latency, boosts data sharing between compute elements, and improves overall system use. These features are key for large-scale AI training and inference in today’s data centers.
“AI infrastructure is no longer built from generic components, it is co-designed from the ground up,” said Patrick Little, President and CEO of SiFive. “By integrating NVLink Fusion with SiFive’s high-performance compute subsystems, we’re enabling customers with an open and customizable CPU platform that pairs seamlessly with NVIDIA’s AI Infrastructure to deliver exceptional efficiency at data center scale.”
“With SiFive, we’re bringing our coherent, high-bandwidth NVLink interconnect to the RISC-V ecosystem,” said Jensen Huang, founder and CEO of NVIDIA. “This provides the flexibility to combine customizable RISC-V CPUs with NVIDIA accelerators to build scalable, power-efficient and specialized AI infrastructure – and move from innovation to deployment faster than ever.”