- Spack usage disclaimer, copyright and trademark notice
- Introduction to SPACK
- Getting Started
- Build Customization
- Technical Support
AMD Toolchain with SPACK
SPACK HPC Applications
Spack is an open source project that offers a package management framework and tool for installing complex scientific software. It is designed to support multiple versions and configurations of a software on many different platforms and environments. Spack supports large supercomputing centers comprising of many users and application teams. They share common installations of software on clusters with complex architectures using libraries without a standard Application Binary Interface (ABI). Many configurations can co-exist on the same system as installing a new version of Spack will not break the existing installations.
Spack provides a simple spec syntax, using which, the users can configure versions and options precisely. It simplifies the job of package authors as the package files are written in Python™. Also, the specs allow them maintain a single file for different builds of a package.
AMD supports the AMD Optimized CPU Compilers and Libraries (AOCC and AOCL ) with Spack packages. AMD also supports Spack packages for commonly used HPC benchmarks and a growing catalogue of scientific, open-source applications with recommended command-line directives using AOCC and AOCL. Each application-specific page details the Spack command line, options, and example instructions to run the application.
For technical support of the tools, benchmarks and applications that AMD offers for Spack building and related inquiries, please refer to technical support.
For information specific to the Spack project and usage, please refer to the Spack documentation.