AMD revealed that it contributed FPGA technology to the U.S. Department of Energy’s newest dedicated science network, a blindingly fast 46 Terabits-per-second architecture called ESnet6.
AMD began with ESnet, headquartered at the Lawrence Berkeley National Laboratory in California, in 2018. The extreme scale packet monitoring system of the current iteration uses AMD Alveo U280 FPGA-based network-attached accelerator cards at the core network switching nodes, according to AMD.
“This will enable high-touch packet processing and help improve the accuracy of network monitoring and management to enhance performance,” according to an AMD statement. “The programmable hardware allows for new capabilities to be added for continuous innovation in the network.”
The semiconductor giant used its open source OpenNIC overlay to provide standard network-interface and host-attachment hardware that allowed the customization of Alveo U280 2x100Gb/s accelerators as network interface cards. ESnet is just one of many academic research groups that have leveraged OpenNIC, AMD said. Also helping to speed development were AMD VitisNetP4 development tools used for compiling the P4 packet processing language to FPGA hardware. AMD Alveo U280 cards, using OpenNIC and VitisNetP4, are being deployed on every node of the ESnet6 network.
The benefits of AMD adaptive computing at demanding network line rates are also applied to the edges of the ESnet6 network. Globally distributed scientific instruments are the primary sources of the ‘big data” science flows that ESnet must support. The Alveo card with OpenNIC can be used for customized data filtering and shaping before transmission over ESnet.
The Alveo card with OpenNIC can be used for load balancing across compute and storage servers working with supercomputing centers. ESnet serves as the “data circulatory system” for the DOE, connecting all of its national laboratories, and tens of thousands of DOE-funded researchers. This includes supercomputer centers like the Frontier exascale system, which was announced this past summer and uses AMD EPYC CPUs and AMD Instinct Accelerators.
“As scientific instruments grow in complexity and supercomputers simulate scientific phenomena at higher resolutions, the science community is facing a growing challenge: data volumes that are increasing exponentially, coupled with the need to move, share, and process this data globally and faster than ever before,” said Barbara Helland, Associate Director of the DOE Office of Science’s Advanced Scientific Computing Research program, in a statement issued by the program. “With ESnet6, DOE researchers are equipped with the most sophisticated technology to help tackle the grand challenges we face today in areas like climate science, clean energy, semiconductor production, microelectronics, the discovery of quantum information science, and more.”