Skip to content

fix: add Pascal CUDA architectures #970

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
nanookclaw wants to merge 1 commit into
base: main
Choose a base branch
from
+1 −0
Open

fix: add Pascal CUDA architectures #970

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
1 change: 1 addition & 0 deletions llamacpp/native/cuda.Dockerfile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,7 @@ RUN echo "-B build \
-DGGML_NATIVE=OFF \
-DGGML_OPENMP=OFF \
-DGGML_CUDA=ON \
-DCMAKE_CUDA_ARCHITECTURES=61;62;70;75;80;86;89;120;121 \

@gemini-code-assist gemini-code-assist Bot Jun 13, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

critical

The CUDA architectures 120 and 121 are invalid. CUDA compute capabilities (architectures) do not go up to 120 or 121; these numbers likely confuse CUDA software versions (12.0, 12.1) with GPU architectures. Passing these invalid values will cause nvcc to fail during compilation with an error like nvcc fatal : Unsupported gpu architecture 'compute_120'.

Additionally, the list is missing Hopper (90), which is a highly important enterprise GPU architecture (e.g., NVIDIA H100).

Please update the list to use valid compute capabilities, such as 61;62;70;75;80;86;89;90.

    -DCMAKE_CUDA_ARCHITECTURES=61;62;70;75;80;86;89;90 \
References
  1. Focus on correctness, safety, and maintainability — in that order. Critical issues include correctness bugs. (link)

-DCMAKE_CUDA_COMPILER=/usr/local/cuda/bin/nvcc \
-DLLAMA_OPENSSL=OFF \
-GNinja \
Expand Down
Loading