enhance custom easyblock for NCCL so it defines $NCCL_HOME in generated module file

[Flamefire]

(created using eb --new-pr)

[Flamefire]

Test report by @Flamefire

Overview of tested easyconfigs (in order)

• SUCCESS NCCL-2.18.3-GCCcore-12.3.0-CUDA-12.1.1.eb
• SUCCESS NCCL-2.12.12-GCCcore-11.3.0-CUDA-11.7.0.eb
• SUCCESS CUDA-12.8.0.eb
• SUCCESS GDRCopy-2.4.4-GCCcore-14.2.0.eb
• SUCCESS UCX-CUDA-1.18.0-GCCcore-14.2.0-CUDA-12.8.0.eb
• SUCCESS NCCL-2.26.6-GCCcore-14.2.0-CUDA-12.8.0.eb

Build succeeded for 6 out of 6 (3 easyconfigs in total)
i7008 - Linux Rocky Linux 8.9 (Green Obsidian), x86_64, AMD EPYC 7702 64-Core Processor (zen2), Python 3.9.18
See https://gist.github.com/Flamefire/e42f268b96b4ac5ce22eb07dede5d9b8 for a full test report.

[boegel]

Test report by @boegel

Overview of tested easyconfigs (in order)

• FAIL (build issue) NCCL-2.18.3-GCCcore-12.3.0-CUDA-12.1.1.eb (partial log available at https://gist.github.com/boegel/58a6b168bb242203f2f411a0fbcfa767)
• SUCCESS NCCL-2.20.5-GCCcore-13.2.0-CUDA-12.4.0.eb

Build succeeded for 1 out of 2 (2 easyconfigs in total)
node3300.joltik.os - Linux RHEL 9.4, x86_64, Intel(R) Xeon(R) Gold 6242 CPU @ 2.80GHz (cascadelake), 1 x NVIDIA Tesla V100-SXM2-32GB, 570.133.20, Python 3.9.18
See https://gist.github.com/boegel/d949e35b478862768dd31456469a22f6 for a full test report.

[Flamefire]

UCX-CUDA/1.14.1-GCCcore-12.3.0-CUDA-12.1.1 is missing?

BTW: LMod coupled with an LLM? Nice!

[boegel]

UCX-CUDA/1.14.1-GCCcore-12.3.0-CUDA-12.1.1 is missing?

BTW: LMod coupled with an LLM? Nice!

It's a proof-of-concept thingie, see https://github.com/boegel/easybuild-llm

[boegel]

Test report by @boegel

Overview of tested easyconfigs (in order)

• SUCCESS NCCL-2.18.3-GCCcore-12.3.0-CUDA-12.1.1.eb
• SUCCESS NCCL-2.20.5-GCCcore-13.2.0-CUDA-12.4.0.eb

Build succeeded for 2 out of 2 (2 easyconfigs in total)
node3300.joltik.os - Linux RHEL 9.4, x86_64, Intel(R) Xeon(R) Gold 6242 CPU @ 2.80GHz (cascadelake), 1 x NVIDIA Tesla V100-SXM2-32GB, 570.133.20, Python 3.9.18
See https://gist.github.com/boegel/dca377dd44f8227eb9c7deb0e699c5b5 for a full test report.

[boegel]

lgtm