Skip to content

Commit

Permalink
Merge branch '23.08-rc1'
Browse files Browse the repository at this point in the history
  • Loading branch information
eugeneswalker committed Aug 31, 2023
2 parents df23bd1 + 1d25f7c commit fbf0b02
Show file tree
Hide file tree
Showing 16 changed files with 31,246 additions and 0 deletions.
107 changes: 107 additions & 0 deletions environments/23.08/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,107 @@
# E4S Release 23.08

August 2023 release of E4S

## Files

* `cuda-x86_64/spack.yaml` -- Model Spack Environment for systems w/ NVIDIA GPUs (x86_64)
* `cuda-ppc64le/spack.yaml` -- Model Spack Environment for systems w/ NVIDIA GPUs (ppc64le)
* `cuda-aarch64/spack.yaml` -- Model Spack Environment for systems w/ NVIDIA GPUs (aarch64)
* `rocm-x86_64/spack.yaml` -- Model Spack Environment for systems w/ AMD GPUs (x86_64)
* `oneapi-x86_64/spack.yaml` -- Model Spack Environment for systems w/ Intel GPUs (x86_64)

Alongside the Spack environments listed above there are `concretize.log` files which show the exact versions and variants of the concretized environment. This includes both the root specs and the entire set of dependencies needed to build them.

*Specs in the Model Spack Environments may be commented out if we encountered deployment issues or the package doesn't offer a fixed version.*


## Spack Version

E4S 23.08 uses Spack branch `e4s-23.08`
* https://github.com/spack/spack
* Branch [e4s-23.08](https://github.com/spack/spack/tree/e4s-23.08)


## Spack Build Cache

* https://cache.e4s.io/23.08 (not intended for web browser access; for use as a `spack mirror`)

```
$> spack mirror add E4S https://cache.e4s.io/23.08
$> spack buildcache keys -it
gpg: key 4345F04B40005581: public key "University of Oregon - E4S" imported
gpg: Total number processed: 1
gpg: imported: 1
gpg: inserting ownertrust of 6
$> spack buildcache list -al
...
```

## Model Environments

* Model environment for NVIDIA GPUs on x86_64 is:
* `cuda-x86_64/spack.yaml`
* For `cuda_arch=80` and `cuda_arch=90`
* Built and tested using:
* `Ubuntu 20.04, GCC 11.4.0`
* `ecpe4s/ubuntu20.04-runner-amd64-gcc-11.4-mpi-base:2023.08.20`

* Model environment for NVIDIA GPUs on ppc64le is:
* `cuda-ppc64le/spack.yaml`
* For `cuda_arch=70`
* Built and tested using:
* `Ubuntu 20.04, GCC 9.4.0`
* `ecpe4s/ubuntu20.04-runner-ppc64-gcc-11.4-mpi-base:2023.08.20`

* Model environment for NVIDIA GPUs on aarch64 is:
* `cuda-aarch64/spack.yaml`
* For `cuda_arch=75`, `cuda_arch=80`, `cuda_arch=90`
* Built and tested using:
* `Ubuntu 20.04, GCC 11.4.0`
* `ecpe4s/ubuntu20.04-runner-arm64-gcc-11.4-mpi-base:2023.08.20`

* Model environment for AMD GPUs on x86_64 is:
* `rocm-x86_64/spack.yaml`
* For `amdgpu_target=gfx908` and `amdgpu_target=gfx90a`
* Built and tested using:
* `Ubuntu 20.04, GCC 11.4.0`
* `ecpe4s/ubuntu20.04-runner-amd64-gcc-11.4-rocm5.4.3-mpi-base:2023.08.20`

* Model environment for Intel GPUs on x86_64 is:
* `oneapi-x86_64/spack.yaml`
* Built and tested using:
* `Ubuntu 20.04, OneAPI 2023.2.1`
* `ecpe4s/ubuntu20.04-runner-amd64-oneapi-2023.2.1-mpi-base:2023.08.20`
* Derived from Intel's official `intel/oneapi-hpckit:2023.2.1-devel-ubuntu20.04` image

## Example output of `spack find`

Here is what `spack find` would show following the installation of the `cuda-x86_64/spack.yaml` with the `cuda_arch=80` specs enabled.


```
root@2cf0ba0183ed:/# spack find -x
-- linux-ubuntu20.04-x86_64 / gcc@11.4.0 ------------------------
adios@1.13.1 caliper@2.10.0 flecsi@2.2.1 heffte@2.3.0 libpressio@0.95.1 nvhpc@23.7 plumed@2.9.0 slepc@3.19.1 turbine@1.3.0
adios2@2.9.1 caliper@2.10.0 flit@2.1.0 hpctoolkit@2023.03.01 libpressio@0.95.1 omega-h@9.34.13 precice@2.5.0 slepc@3.19.1 umap@2.1.0
adios2@2.9.1 chai@2022.03.0 flux-core@0.53.0 hpctoolkit@2023.03.01 libquo@1.3.1 omega-h@9.34.13 pruners-ninja@1.0.1 stc@0.9.0 umpire@2022.03.1
alquimia@1.0.10 chai@2022.03.0 flux-core@0.53.0 hpx@1.9.1 libunwind@1.6.2 openfoam@2306 pumi@2.2.7 strumpack@7.1.3 umpire@2022.03.1
aml@0.2.1 charliecloud@0.33 fortrilinos@2.3.0 hpx@1.9.1 loki@0.1.7 openmpi@4.1.5 py-cinemasci@1.3 strumpack@7.1.3 unifyfs@1.1
amrex@23.08 conduit@0.8.8 gasnet@2023.3.0 hypre@2.29.0 magma@2.7.1 openpmd-api@0.15.1 py-h5py@3.8.0 sundials@6.5.1 upcxx@2023.3.0
amrex@23.08 cuda@11.8.0 ginkgo@1.6.0 hypre@2.29.0 mercury@2.3.0 papi@6.0.0.1 py-jupyterhub@1.4.1 sundials@6.5.1 upcxx@2023.3.0
arborx@1.4.1 cusz@0.3.1 ginkgo@1.6.0 kokkos@4.1.00 metall@0.25 papi@6.0.0.1 py-libensemble@0.10.2 superlu@5.3.0 variorum@0.6.0
arborx@1.4.1 darshan-runtime@3.4.4 globalarrays@5.8.2 kokkos@4.1.00 mfem@4.5.2 papyrus@1.0.2 py-petsc4py@3.19.4 superlu-dist@8.1.2 veloc@1.6
argobots@1.1 darshan-util@3.4.4 gmp@6.2.1 kokkos-kernels@4.0.00 mfem@4.5.2 parallel-netcdf@1.12.3 py-torch@2.0.1 superlu-dist@8.1.2 visit@3.3.3
ascent@0.9.1 datatransferkit@3.1-rc3 gotcha@1.0.4 kokkos-kernels@4.0.00 mgard@2023-03-31 paraview@5.11.1 py-warpx@23.08 swig@4.0.2-fortran vtk-m@1.9.0
axom@0.7.0 dealii@9.4.2 gptune@4.0.0 lammps@20230802 mgard@2023-03-31 paraview@5.11.1 qthreads@1.18 sz@2.1.12.5 vtk-m@2.0.0
bolt@2.0 dealii@9.4.2 h5bench@1.4 lammps@20230802 mpark-variant@1.4.0 parsec@3.0.2209 quantum-espresso@7.2 sz3@3.1.7 wannier90@3.1.0
boost@1.79.0 dyninst@12.3.0 hdf5@1.12.2 lbann@0.102 mpich@4.1.2 parsec@3.0.2209 raja@2022.10.4 tasmanian@7.9 xyce@7.6.0
bricks@r0.1 ecp-data-vis-sdk@1.0 hdf5@1.12.2 legion@23.06.0 mpifileutils@0.11.1 pdt@3.25.1 raja@2022.10.4 tasmanian@7.9 zfp@0.5.5
bricks@r0.1 ecp-data-vis-sdk@1.0 hdf5-vol-async@1.7 legion@23.06.0 nccmp@1.9.1.0 petsc@3.19.4 rempi@1.1.0 tau@2.32.1 zfp@0.5.5
butterflypack@2.2.2 exaworks@0.1.0 hdf5-vol-cache@v1.1 libcatalyst@2.0.0-rc3 nco@5.1.6 petsc@3.19.4 scr@3.0.1 tau@2.32.1
cabana@0.5.0 faodel@1.2108.1 hdf5-vol-log@1.4.0 libcatalyst@2.0.0-rc3 netlib-scalapack@2.2.0 phist@1.11.2 slate@2022.07.00 trilinos@14.4.0
cabana@0.5.0 flecsi@2.2.1 heffte@2.3.0 libnrm@0.1.0 nrm@0.1.0 plasma@22.9.29 slate@2022.07.00 trilinos@14.4.0
==> 168 installed packages
```
Loading

0 comments on commit fbf0b02

Please sign in to comment.