By combining wide, out-of-order cores with scalable coherent fabrics and advanced memory hierarchies, SiFive enables cloud providers and system vendors to tailor their CPUs to workload requirements while also maintaining the software portability that comes with the RISC-V ecosystem.
With the addition of NVLink, SiFive can connect directly to Nvidia GPUs and accelerators using a coherent, high-bandwidth interconnect to reduce latency, share data more efficiently, and improve overall system utilisation for large-scale AI deployments.
“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 customisable RISC-V CPUs with NVIDIA accelerators to build scalable, power-efficient and specialised AI infrastructure – and move from innovation to deployment faster than ever.”
NVLink Fusion provides a unified interconnect fabric capable of coherently linking CPUs, GPUs, and custom accelerators at extremely high bandwidth. For SiFive customers, this means faster integration cycles, simpler system design, and the ability to scale AI performance without being constrained by traditional CPU-to-accelerator bottlenecks.
