Next-Gen Networking for Heterogeneous Compute & Storage

The slowing down of Moore’s-law coupled with the explosion of new storage and compute intensive applications like deep learning, AI and IoT have created a severe challenge for data center networking. In the past, these problems were the domain of High Performance Computing but now enterprise, edge and cloud segments are all impacted. Incumbent technologies like Ethernet, Fibre Channel and Infiniband are unable to deliver the bandwidth and latency performance especially for smaller clusters. GigaIO’s FabreX network is a PCIe standards-based solution that addresses these challenges. In addition to providing unparalleled latency and bandwidth performance, FabreX is capable of natively supporting NVMe-oF and GDR protocols, removing the extra overhead associated with transferring data over another transport. We will present our S/W and H/W architecture for NVME-oF transport support and disaggregated Xilinx Alveo cards running a machine learning suite. Measured data from internal testing and from San Diego Supercomputer Center (SDSC) will be presented to demonstrate the performance and efficiency benefits of our solution.