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.

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

[ROCM] Raise device memory cap for parallel GPU execution to 5GB #2840

Open
wants to merge 1 commit into
base: develop-upstream
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
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
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@ function is_absolute {
[[ "$1" = /* ]] || [[ "$1" =~ ^[a-zA-Z]:[/\\].* ]]
}

export TF_PER_DEVICE_MEMORY_LIMIT_MB=${TF_PER_DEVICE_MEMORY_LIMIT_MB:-4096}
export TF_PER_DEVICE_MEMORY_LIMIT_MB=${TF_PER_DEVICE_MEMORY_LIMIT_MB:-5120}

# *******************************************************************
# This section of the script is needed to
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -1445,8 +1445,6 @@ CHECK: tt.store
// Reproducer from b/384110192.
TEST_F(TritonEmitterTest,
FusionWithOutputContainingMoreThanInt32MaxElementsExecutesCorrectly) {
GTEST_SKIP() << "TODO(rocm): Weekly-sync 25-01-21: Skip Int32 max elements "
"issue with triton.";
// The point here is to check the output of the Triton fusion. The `slice` op
// at the end is inserted to allow the comparison of output to run in a
// reasonable amount of time, and has been proven to still correctly capture
Expand Down