HPE Cray Supercomputing Operating System Software

Improving Workload Performance

HPE Cray OS features CPU assignment which constrains OS operations to specific Linux hardware threads eliminating noise from the remaining CPUs so that applications have dedicated CPUs for a better workload performance.

Additional Huge Page sizes we added deliver less latency by avoiding address translation cache misses in limited-sized network address translation hardware - important for applications with large memory footprints and sparse reference patterns.

XPMEM – enables enhanced memory sharing between processes to increase on-node MPI performance.

The software controls turbo boost to provide predictable, consistent performance of the system.

Going Beyond Standard Linux Capabilities

HPE Cray OS offers fine-grained monitoring of power usage of the running applications so users can decide whether they want to contain or exploit available power resources and make sure their system runs efficiently.

Advanced Kernel Debugging extends the crash capabilities of the community Linux kernel debugger supporting diskless nodes and huge pages for large scale systems. Node Memory Dump enables administrators to select which nodes to dump in the event of a system crash.

The software includes native clients that enable access to Lustre file systems. Data Virtualization Service (DVS) provides compute nodes with scalable access to file systems mounted on service nodes and also enables lightweight access to NFS, IBM Spectrum and Lustre file systems.

Leverage Cray Supercomputing Lineage

HPE Cray Operating System was developed and enhanced over several decades to address specific needs of HPC customers.

This customized (not custom!) OS does not alter the ability to run standard Linux applications, but rather enhance it for performance, scale, and reliability.

HPE Cray OS is fully supported by HPE Pointnext Services.