Spack

I’ll introduce Spack by quoting the first paragraph of their documentation:

Spack is a package manager for supercomputers, Linux, and macOS. It makes installing scientific software easy. Spack isn’t tied to a particular language; you can build a software stack in Python or R, link to libraries written in C, C++, or Fortran, and easily swap compilers or target specific microarchitectures.

So what does this mean in practice? You’ve already learnt about build systems like Autotools and CMake, and may be familiar with module systems as used on ARC. You may also have worked through the section on containers and wondered if there was an easier way of building a potentially deep stack of software for a particular project.

You may have already tried Conda and found it wonderful, but for the fact that the piece of software you want isn’t built with the options you wanted, so you can’t use it. Or maybe you just can’t make the conda installed version work with the MPI setup on the HPC cluster you’re using.

In all these cases, Spack has the potential to make your life a little bit easier, and the learning curve is such that you can incrementally learn about Spack, and find it useful almost straight away.

As such, I’m going to introduce Spack with a series of examples, that work you way up what you can do with Spack, without going too far down into the details. If you do want to learn more about Spack, the official Spack documentation is mostly excellent.

• Installing Spack

• Building a test piece of software

• How you can tell it about bits of software you already have

• An advanced software build example

• Building containers with Spack

• Using Spack to Replace Homebrew/Conda

• Writing a Spack recipe to build a bit of software not already handled by Spack

• Using environment modules with Spack