-
Couldn't load subscription status.
- Fork 16
Add NVHPC/21.11 and CUDA/11.5.1 #1392
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you file this against merge-upstream, too?
|
Retest this please Jenkins. |
2 similar comments
|
Retest this please Jenkins. |
|
Retest this please Jenkins. |
|
I believe the failures are because the new NVC++ with its default target architecture ( I'm not quite sure what's going on, as suggests it is auto-detecting while compiles successfully. I guess we need to report this to the NVIDIA forums. I also see in the generated which looks like it has too much irrelevant stuff in it. |
75606e6 to
b71ea24
Compare
|
Note that the release notes say 11.5.1 is included, so I think we should install that one. Hopefully Spack will learn how to express that |
|
@pramodk I can't request a review from you because you're the original author 😅 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Overall LGTM if everything is working in production :)
* Add + deploy NVIDIA HPC SDK 21.11. * Add + deploy CUDA 11.5.1 to match it. * Do not define CMAKE_C_FLAGS in `coreneuron+nmodl`. * Run `makelocalrc` for `nvhpc` in a cleaner environment. * Add `-mno-abm` to [Core]NEURON recipes for `[email protected]`. Co-authored-by: Olli Lupton <[email protected]>
Presumably this was working before because our nvhpc localrc files accidentally included CUDA include directories before BlueBrain/spack#1392.
* Add + deploy NVIDIA HPC SDK 21.11. * Add + deploy CUDA 11.5.1 to match it. * Do not define CMAKE_C_FLAGS in `coreneuron+nmodl`. * Run `makelocalrc` for `nvhpc` in a cleaner environment. * Add `-mno-abm` to [Core]NEURON recipes for `[email protected]`. Co-authored-by: Olli Lupton <[email protected]>
* Cherry-pick of #1392. * Add + deploy NVIDIA HPC SDK 21.11. * Add + deploy CUDA 11.5.1 to match it. * Do not define CMAKE_C_FLAGS in `coreneuron+nmodl`. * Run `makelocalrc` for `nvhpc` in a cleaner environment. * Use GCC 11.2.0 for NVHPC instead of GCC 9.4.0. * Add `-mno-abm` to [Core]NEURON recipes for `[email protected]`. Co-authored-by: Pramod Kumbhar <[email protected]>
Presumably this was working before because our nvhpc localrc files accidentally included CUDA include directories before BlueBrain/spack#1392.
-mno-abmto 21.11 NVIDIA compilers to avoid an issue with the__ABM__macro and Random123 (https://forums.developer.nvidia.com/t/21-11-tp-behaviour/198115)localrcgeneration for NVIDIA compilers to run in a cleaner environment. Previously this ran after compiler configuration for LLVM, andmodule load llvm ... module unload llvmleft LLVM's autoloaded dependencies,pythonandcuda, in the environment, and those paths made their way into thelocalrcfile.CMAKE_C_FLAGSfor CoreNEURON + NMODL builds that contain no C code.deploy/set-compiler-flags.py-- cc: @matz-e who was interested in tweaking it.