HPE Supercomputing Programming Software 1.0.0

HPE Supercomputing Programming Software (SUPS) suite offers programmers a comprehensive set of tools for developing, porting, debugging, and tuning applications. The programming environment simplifies the transition to new hardware architectures and configurations by automatically applying optimizations on HPC applications that use existing programming models with a simple recompile.

Quick Links:

If you have any suggestions or find any errors in our documentation, please join our Slack Workspace and start a discussion in our #hpe-cray-programming-environment channel.

Note: Online documentation for OpenSHMEMX and dsmml has moved and can be found here.

HPE Cray Compiling Environment

Our Fortran compiler is designed to help extract maximum performance from the systems regardless of the underlying architecture supporting ARM and x86-64 (Intel and AMD) processors, as well as AMD and NVIDIA accelerators.

The compiler identifies regions of computation that are either sequential scalar or vector parallel, and automatically exploits these capabilities for the targeted system.

The compiler gives programmers optimization feedback with an annotated listing of source code for easier application tuning. Our compiler integrates with debuggers and performance tools in the HPE Supercomputing Programming Software suite—enhancing each other’s capability to generate correct and performant code.

The suite also includes integration with GNU, Intel, AMD, and NVIDIA programming environments, so developers can choose between multiple compilers and still use the libraries, debuggers, and performance analysis tools included in our suite to help optimize application performance.

We focus on standards compliance for code safety, application portability, and investment protection. Our compiler supports standard Fortran and standard programming models such as OpenMP and OpenACC.

HPE Cray Message Passing Toolkit

CMPT is a collection of libraries that provide portable, efficient, and flexible mechanisms for performing data transfers between parallel processes. It comprises HPE Cray MPI, and HPE Cray PMI.

HPE Cray MPI is an MPICH ABI compatible library tuned for Intel, AMD, and ARM CPUs as well as AMD and NVIDIA GPUs. It is a highly scalable implementation, customized for low latency and high bandwidth, both on-node and off-node, for point-to-point and collective communications. It is also highly optimized and tuned for the HPE Slingshot network architecture. Strategic optimizations for MPI I/O, MPI_THREAD_MULTIPLE, remote memory access (RMA), and integration with the performance analysis tools in the suite contribute to deliver ideal application performance for today’s HPC codes.

Generated Oct 13, 2026