Skip to content

Use 256-thread blocks for jagged dense-output kernel (#5848) #5848

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

Open
MericGit wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Use 256-thread blocks for jagged dense-output kernel (#5848) #5848

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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: 6 additions & 0 deletions fbgemm_gpu/src/jagged_tensor_ops/common.cuh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -224,7 +224,13 @@ inline std::tuple<dim3, dim3, StackArray<int64_t>> check_shape_and_partition_(

const int threads_x =
inner_dense_size >= kWarpSize / 2 ? kWarpSize : inner_dense_size;
#ifndef USE_ROCM
const int threads_y = kMaxThreads / kWarpSize;
#else
// AMD: ~256-thread blocks improve wavefront packing for common D=1 shapes.
constexpr int kTargetBlockThreads = 256;
const int threads_y = kTargetBlockThreads / threads_x;
#endif
const dim3 blocks(
div_round_up(outer_dense_size * jagged_folded_size, threads_y));

Expand Down
Loading