Skip to content

Fallback to default CUDA_ARCH if nvidia-smi fails #4624

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 2 commits into from
Aug 21, 2025
Merged

Fallback to default CUDA_ARCH if nvidia-smi fails #4624

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7d1b926
Add fallback value for `CUDA_ARCH` if nvidia-smi fails
hafeoz Aug 21, 2025
9470add
Simplify `CUDA_ARCH` logic
hafeoz Aug 21, 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
10 changes: 2 additions & 8 deletions Tools/GNUMake/sites/Make.unknown
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,17 +94,11 @@ ifeq ($(USE_CUDA),TRUE)
cuda_device_query_result := $(shell $(CUDA_HOME)/extras/demo_suite/deviceQuery | grep "CUDA Capability")
else
# try nvidia-smi, as deviceQuery is no longer distributed with CUDA 12 and up
cuda_device_query_result := $(shell nvidia-smi --query-gpu=compute_cap --format=csv,noheader)
cuda_device_query_result := $(shell nvidia-smi --query-gpu=compute_cap --format=csv,noheader || echo $(CUDA_ARCH))
endif
ifneq ($(cuda_device_query_result),)
TEMP_CUDA_ARCH = $(lastword $(cuda_device_query_result))
# CUDA_ARCH = $(subst .,,$(TEMP_CUDA_ARCH))
# We do this because pgfortran does not support all cuda arches.
ifeq ($(TEMP_CUDA_ARCH),3.5)
CUDA_ARCH = 35
else
CUDA_ARCH = $(firstword $(subst ., ,$(TEMP_CUDA_ARCH)))0
endif
CUDA_ARCH = $(subst .,,$(TEMP_CUDA_ARCH))
endif
endif
endif
Expand Down
Loading