Are your high-performance computing (HPC) applications memory-bound? The HPE Apollo 35 System provides a high density AMD EPYC™ compute solution with exceptional memory bandwidth and capacity per socket. Offering four servers in 2U, with up to 32 cores and 1 TB DDR4 memory per socket, the HPE Apollo 35 System is well suited for memory bandwidth or capacity bound HPC workloads, such as computational fluid dynamics, weather simulation, and oil and gas exploration. With high-performance core density per server and support for high-bandwidth network interfaces, the HPE Apollo 35 System delivers excellent HPC performance at both a node and rack level.
High Throughput Performance for Memory-Bound HPC Applications
The HPE Apollo 35 System has exceptional memory bandwidth, with eight DDR4 memory channels, and as demonstrated by the Stream benchmark.1
Up to 32 cores per socket for increased parallel processing power.
World record floating point performance as demonstrated by the SPEC SPU benchmark.2
Improves Total Cost of Ownership Through Efficient Design
The HPE Apollo 35 System provides twice the density of 1U servers - four 1U servers in a 2U chassis.
Consolidates infrastructure through higher core count processors, saving on power, cooling, and material costs.
HPE Apollo 35 System provides excellent density with 168 sockets and 5376 cores per rack to reduce datacenter sprawl.
Superb for High Performance Computing Environments
The HPE Apollo 35 System is qualified by major computer-aided engineering (CAE) vendors, such as ANSYS and Altair, to validate and enhance application performance and stability.
Supports HPE InfiniBand EDR Adapters for low latency communication and high bandwidth between nodes.
Certified with HPE Performance Cluster Manager for fast system setup, comprehensive hardware monitoring, image management, software updates, and power management.3
- 1. https://www.amd.com/system/files/2017-06/AMD-EPYC-SoC-Delivers-Exceptional-Results.pdf
- 2. https://www.amd.com/system/files/2018-03/AMD-SoC-Sets-World-Records-SPEC-CPU-2017.pdf
- 3. Support for HPE Performance Cluster Manager will begin in December 2018.
- AMD EPYC is a trademark of Advanced Micro Devices, Inc. in the United States and/or other jurisdictions. All other third-party marks are property of their respective owners.