qmcpack icon indicating copy to clipboard operation
qmcpack copied to clipboard
verified publisher icon QMCPACK

Add offload build support in Spack

Open ye-luo opened this issue 3 years ago • 4 comments

Is your feature request related to a problem? Please describe. Add offload build support in Spack

ye-luo avatar Nov 10 '22 20:11 ye-luo

Just noticed that e4s was doing some nightly tests. https://gitlab.e4s.io/uo-public/qmcpack/-/pipelines Although they used a node with a100 but no GPU feature is turned on.

ye-luo avatar Mar 27 '23 17:03 ye-luo

Once https://github.com/spack/spack/pull/36675 gets in we should be able to work on

  1. Updating +cuda variant if needed
  2. add +rocm variant with the rocm package
  3. add +offload variant

ye-luo avatar Apr 06 '23 16:04 ye-luo

~~https://github.com/spack/spack/pull/36675~~ has been merged. If we enable offload features of qmcpack in spack, we can test the integration. When using llvm compilers, add +libomptarget requirement.

ye-luo avatar May 22 '23 18:05 ye-luo

Thanks for the update. Another reason to have the e4s build run with offload is that it will highlight the current cuda <=11.2 dependency / ptx bug/problem with offload.

prckent avatar May 22 '23 18:05 prckent