{numlib}[NVHPC/24.9-CUDA-12.6.0] FlexiBLAS v3.4.4, OpenBLAS v0.3.27 w/ fb#21635
{numlib}[NVHPC/24.9-CUDA-12.6.0] FlexiBLAS v3.4.4, OpenBLAS v0.3.27 w/ fb#21635sassy-crick wants to merge 3 commits intoeasybuilders:developfrom
Conversation
…AS-0.3.27-NVHPC-24.9-CUDA-12.6.0.eb, ScaLAPACK-2.2.0-nvompi-2024.9-fb.eb
Updated software
|
|
Test report for EasyBuild/5.0.0. can be found here |
|
@boegelbot please test @ jsc-zen3 |
|
@sassy-crick: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de PR test command '
Test results coming soon (I hope)... Details- notification for comment with ID 2805398045 processed Message to humans: this is just bookkeeping information for me, |
|
Test report by @boegelbot |
…asyconfigs into 20241014123704_new_pr_FlexiBLAS344
|
@boegelbot please test @ jsc-zen3 |
|
@sassy-crick: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de PR test command '
Test results coming soon (I hope)... Details- notification for comment with ID 2997635880 processed Message to humans: this is just bookkeeping information for me, |
|
Test report by @boegelbot |
|
Now, that is something I don't really understand: Why was that all working a few months ago with the |
|
FlexiBLAS and NVHPC can be a bit of a nightmare. With this exact combination, we saw: on aarch64. The warning itself is concerning enough. The failure because of So even if the build works, we really need to carefully investigate if the build is actually doing what we expect. |
|
With the new NVHPC toolchain being introduced in EasyBuild v5.2.0, I'm not sure if we should promote the old NVHPC toolchains anymore, even if this is targeting an old toolchain generation... |
Agreed. That PR stems from a time before we had converged on how to set up |
(created using
eb --new-pr)