Skip to content

fix: add SM90 guard for FP8 Blockscale GEMM #3575

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

Merged
merged 3 commits into from
Apr 16, 2025
Merged

fix: add SM90 guard for FP8 Blockscale GEMM #3575

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
fadb1a8
fix: add SM90 guard for FP8 Blockscale GEMM
lucifer1004 Apr 15, 2025
892b5a7
fix: add SM90 guard for FP8 Blockscale GEMM
lucifer1004 Apr 15, 2025
b47198a
Merge branch 'main' into user/zihuaw/fp8_blockscale_gemm_hopper_guard
litaotju Apr 16, 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
9 changes: 9 additions & 0 deletions cpp/include/tensorrt_llm/deep_gemm/compiler.cuh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -307,6 +307,15 @@ public:
uint32_t const block_k, uint32_t const num_groups, uint32_t const num_stages, uint32_t const num_tma_multicast,
deep_gemm::GemmType const gemm_type)
{
int sm_version = tensorrt_llm::common::getSMVersion();
if (sm_version != 90)
{
TLLM_THROW(
"DeepGEMM only supports Hopper (SM90) architectures, but current device compute "
"capability is %d.",
sm_version);
}

// Build signature - simplified, no MD5 calculation
std::string name = "gemm_" + std::to_string(shape_n) + "_" + std::to_string(shape_k) + "_"
+ std::to_string(block_m) + "_" + std::to_string(block_n) + "_" + std::to_string(block_k) + "_"
Expand Down
Loading