Open Compute Project and UALink Consortium Collaborate on AI Interconnects

The Open Compute Project Foundation and UALink Consortium have announced a new collaboration to optimize scale-up interconnect performance for AI and High-Performance Computing (HPC) clusters. Announced in a press release, this partnership aims to integrate UALink's scale-up AI interconnect technology into OCP Community-delivered AI clusters, providing the necessary high-bandwidth, low-latency, and low-power connectivity for advanced AI training and inference.

The collaboration will focus on aligning OCP's community-led infrastructure development with UALink's interconnect innovations. This integration is expected to enhance the performance of large-scale AI clusters by addressing critical challenges in interconnect bandwidth and scalability. The alliance will leverage the expertise of both organizations to advance scale-up AI interconnect performance, following the recent release of the UALink 1.0 Specification.

The UALink Consortium, which includes industry leaders like Alibaba, AMD, and Microsoft, is dedicated to developing open industry standards for high-performance accelerated compute scale-up interconnects tailored for AI and HPC workloads. This collaboration with the OCP Community is set to accelerate the adoption of UALink's innovations into complete systems, delivering transformative performance for AI markets.