Are you looking for more choices and flexibility for your high performance computing (HPC) cluster system architecture with the quality and support of a Tier 1 server supplier?
The HPE Apollo 70 System is designed to bring real change and choice to the HPC market by delivering outstanding performance our first arm-based purpose built platform. Supporting the Marvell ThunderX2® Arm processor, the dual-socket server nodes deliver the high performance, high core count, and high memory bandwidth ideal for memory intensive HPC workloads. The HPE software innovation and Linux® operating system flexibility complement the hardware innovation to deliver increased application performance, system reliability, and management efficiency for the most demanding HPC environments including manufacturing, health and life sciences, oil and gas and geosciences.
Bringing Change and Choice to the High Performance Computing Community
The HPE Apollo 70 System features the high performance Marvell ThunderX2 Arm processor, enabling straightforward access to Arm architecture based high performance computing (HPC).
Using the high performance Marvell ThunderX2 processors with up to 32 cores and 8 DDR4 memory channels, the HPE Apollo 70 System supports higher core counts and 33% more memory bandwidth than typical industry standard dual socket server platforms.
Choice of two compute tray offerings enables flexibility in choice of either higher number of nodes per rack, or PCIe expansion capabilities, or a combination of both.
Designed from the Ground Up for HPC Cluster Deployments
The HPE Apollo 70 System is a dense “4 in 2U” chassis system architecture that supports up to 4 dual socket servers in a 2U chassis and enables up to 80 nodes, allowing over 5,000 Arm cores in a 42U rack.
The Marvell ThunderX2 processor system-on-a-chip delivers increased density and power efficiency which is key to building larger cluster installations. Each server node feature built in high speed (100 Gb/s) Mellanox® ConnectX®-5 InfiniBand or Ethernet connectivity for cluster communication.
Reduce the time and resources spent administering your system using HPE Performance Cluster Manager software which offers all functionalities you need to manage your cluster all day everyday including system setup, hardware and health management, image management, power management and more.
Quality and Support Expected from Hewlett Packard Enterprise Design and Manufacture
The HPE Apollo 70 System leverages the proven, high density, cost effective HPE Apollo 2000 System architecture which has been a leading HPC cluster building block for multiple generations.
Combined with the Hewlett Packard Enterprise HPC robust ecosystem of software developed by HPE, open source, and leading commercial software to deliver a choice of operating systems, development tools, workload management, and more for Arm-based solutions.
HPE Pointnext support provides a comprehensive portfolio including Advisory, Professional, and Operational Services. HPE Pointnext also offers a flexible infrastructure consumption model which can bring the economics and flexibility of cloud to your on premises data center.
- Arm is a registered trademark of Arm Limited. Linux is the registered trademark of Linus Torvalds in the U.S. and other countries. All other third-party trademark(s) is/are property of their respective owner(s). Mellanox® is a registered trademark of Mellanox Corporation in the U.S. and other countries.
HPE Apollo Systems
- HPE Apollo Systems
- Evolving the Supercomputer: HPE Apollo 70 is a new model for innovation in the HPC market
- 3D Demo