[Bugfix] prioritize architecture capability before DeepGEMM availability check (#58073)

Signed-off-by: Tony Lin <tony.lin@intel.com>
This commit is contained in:
Tony Lin
2026-09-22 11:38:04 -04:00
committed by GitHub
parent feb87b1934
commit c42f5285ec
+1 -1
View File
@@ -112,7 +112,7 @@ def is_deep_gemm_supported() -> bool:
Currently, only Hopper and Blackwell GPUs are supported.
"""
is_supported_arch = current_platform.support_deep_gemm()
return envs.VLLM_USE_DEEP_GEMM and has_deep_gemm() and is_supported_arch
return is_supported_arch and envs.VLLM_USE_DEEP_GEMM and has_deep_gemm()
@functools.cache