Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

benchmark scripts: basic separation for alps vcluster, number of cores, gpu defaults, ... #1254

Draft
wants to merge 5 commits into
base: master
Choose a base branch
from
Draft

benchmark scripts: basic separation for alps vcluster, number of cores, gpu defaults, ... #1254

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
819f1d1
rename to alps-gh200 and create a copy for each vcluster that inherit…
albestro Dec 20, 2024
f90db2b
limit to 64 threads (slurm allocation, not just pika)
albestro Dec 20, 2024
1d327a7
WIP: is the default the right one or is it better 31+1?
albestro Dec 20, 2024
1d30d37
WIP: change some gpu "defaults" parameter
albestro Dec 20, 2024
f9d3bb0
minor fix: miniapp bt_red2band had the wrong name
albestro Jan 9, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions miniapp/miniapp_bt_reduction_to_band.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -102,7 +102,7 @@ struct Options
};
}

struct BacktransformBandToTridiagMiniapp {
struct BacktransformReductionToBandMiniapp {
template <Backend backend, typename T>
static void run(const Options& opts) {
using MatrixMirrorType = MatrixMirror<T, DefaultDevice_v<backend>, Device::CPU>;
Expand Down Expand Up @@ -237,7 +237,7 @@ int pika_main(pika::program_options::variables_map& vm) {
dlaf::ScopedInitializer init(vm);
const Options opts(vm);

dlaf::miniapp::dispatchMiniapp<BacktransformBandToTridiagMiniapp>(opts);
dlaf::miniapp::dispatchMiniapp<BacktransformReductionToBandMiniapp>(opts);

return EXIT_SUCCESS;
}
Expand All @@ -248,7 +248,7 @@ int main(int argc, char** argv) {

// options
using namespace pika::program_options;
options_description desc_commandline("Usage: miniapp_bt_band_to_tridiag [options]");
options_description desc_commandline("Usage: miniapp_bt_reduction_to_band [options]");
desc_commandline.add(dlaf::miniapp::getMiniappOptionsDescription());
desc_commandline.add(dlaf::getOptionsDescription());

Expand Down
13 changes: 11 additions & 2 deletions scripts/systems.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,8 +198,8 @@
""",
}

cscs["santis"] = {
"Cores": 288,
cscs["alps-gh200"] = {
"Cores": 256,
"Threads per core": 1,
"Allowed rpns": [4],
"Multiple rpn in same job": True,
Expand All @@ -217,12 +217,19 @@
#SBATCH --no-requeue

# Env

export FI_MR_CACHE_MONITOR=disabled
export MPICH_GPU_SUPPORT_ENABLED=1
export MIMALLOC_EAGER_COMMIT_DELAY=0
export MIMALLOC_ALLOW_LARGE_OS_PAGES=1

export PIKA_MPI_COMPLETION_MODE=31
export PIKA_MPI_ENABLE_POOL=1

export DLAF_BT_BAND_TO_TRIDIAG_HH_APPLY_GROUP_SIZE=128
export DLAF_UMPIRE_DEVICE_MEMORY_POOL_ALIGNMENT_BYTES=$((1 << 21)) # 2 MiB, large page size
export DLAF_RED2BAND_PANEL_NWORKERS=50
export DLAF_TFACTOR_NWORKERS=4

# Debug
module list &> modules_{bs_name}.txt
Expand All @@ -232,6 +239,8 @@
""",
}

cscs["alps-todi"] = cscs["alps-gh200"].copy()

csc = {}

csc["lumi-cpu"] = {
Expand Down
Loading