Arm has announced that the Neoverse V2, codenamed “Demeter”, will be added to its Arm Neoverse roadmap. Neoverse V2, the latest Arm core, is designed to deliver leading-edge per-thread performance for cloud, hyperscale, and HPC workloads, the foundation of its next-generation infrastructure solutions.
Arm said it launched the Neoverse V2 platform to further improve cloud workload performance without increasing power and area for hyperscale and HPC customers. Equipped with the latest V-Series cores and widely deployed Arm CMN-700 mesh interconnect technology, coupled with the introduction of Armv9 architecture security enhancements, it will deliver market-leading performance for cloud and HPC workloads.
At present, many Arm partners have invested in the design based on the Neoverse V2 platform. One of them is Nvidia, which will utilize the Neoverse V2 platform as the computing base for its Grace data center CPUs, and combines the power efficiency of the Neoverse V2 platform with LPDDR5X memory to provide twice the performance per watt than traditional architecture servers.
The Neoverse E series is currently being widely used in data plane processing, 5G RAN, edge networks, and accelerators. To support its diversity, Arm also offers the Neoverse E2 platform, combining the Arm Cortex-A510 core with the scalable Neoverse CMN-700 and N2 system backplane, allowing the use of the best cloud technologies in constrained applications, including scalable core count range, SystemReady compatibility, and PCIe, CXL, IO and interfaces.
Arm said that the next-generation N-series products are under development and will be available to partners in 2023, which will further improve both performance and efficiency compared to the existing Neoverse N2 platform.