Skip to content

[Bugfix] Add 12.1 to CUDA_SUPPORTED_ARCHS for CUDA-13 builds (sm_121a/GB10) #43003

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

Closed
ubehera wants to merge 1 commit into from
Closed

[Bugfix] Add 12.1 to CUDA_SUPPORTED_ARCHS for CUDA-13 builds (sm_121a/GB10) #43003

Show file tree
Hide file tree
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
13 changes: 12 additions & 1 deletion CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -101,7 +101,18 @@ if(DEFINED CMAKE_CUDA_COMPILER_VERSION AND
# starting from CUDA 12.9 and Blackwell (10.0), we use family-specific targets (10.0f, 12.0f, etc)
# to support the whole generation without specifying all sub-architectures
# see: https://developer.nvidia.com/blog/nvidia-blackwell-and-nvidia-cuda-12-9-introduce-family-specific-architecture-features/
set(CUDA_SUPPORTED_ARCHS "7.5;8.0;8.6;8.7;8.9;9.0;10.0;11.0;12.0")
# 12.1 is listed explicitly because the FP8 Marlin kernels (MARLIN_FP8_ARCHS,
# MARLIN_MOE_FP8_ARCHS) request plain 12.1 (no suffix) and need it to survive the
# intersection at line 185 below. Without 12.1 here, sm_121a is dropped from
# CUDA_ARCHS and no sm_121a Marlin SASS is generated; the matching runtime
# arch check in csrc/quantization/marlin/marlin.cu is also relaxed to allow
# the whole sm_12x family (matching the existing MoE path in
# csrc/moe/marlin_moe_wna16/ops.cu). Without both, FP8 quant inference on
# GB10 / DGX Spark (sm_121a) fails with cudaErrorNoKernelImageForDevice
# (binary missing) or a TORCH_CHECK failure (runtime guard). The 12.0f
# family targets used by MLA/SCALED_MM/FP4/CUTLASS_MOE_DATA already cover
# sm_121 via family-compatibility and are unaffected by this change.
set(CUDA_SUPPORTED_ARCHS "7.5;8.0;8.6;8.7;8.9;9.0;10.0;11.0;12.0;12.1")
Comment thread
ubehera marked this conversation as resolved.
elseif(DEFINED CMAKE_CUDA_COMPILER_VERSION AND
CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 12.8)
set(CUDA_SUPPORTED_ARCHS "7.5;8.0;8.6;8.7;8.9;9.0;10.0;10.1;10.3;12.0;12.1")
Expand Down
6 changes: 4 additions & 2 deletions csrc/quantization/marlin/marlin.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -400,10 +400,12 @@ void marlin_mm(const void* A, const void* B, void* C, void* C_tmp, void* b_bias,
"Turing only support FP16 or INT8 activation.");
}
if (a_type == vllm::kFE4M3fn) {
TORCH_CHECK(major_capability * 10 + minor_capability >= 89,
"FP8 only support Ada Lovelace or newer GPUs.");
TORCH_CHECK(
major_capability * 10 + minor_capability == 89 ||
major_capability * 10 + minor_capability == 120,
"Marlin W4A8-FP8 only support SM89 or SM120 device (It is slower than "
major_capability == 12,
"Marlin W4A8-FP8 only support SM89 or SM12x device (It is slower than "
"Marlin W4A16 on other devices).");
}

Expand Down
Loading