Compiling PyTorch CUDA Extensions in Docker Without a GPU

Mon, 17 Aug 2026 14:25:36 +0800

One detail that is easy to forget when building CUDA containers is that the GPU itself is normally not available during docker build. The compiler and CUDA toolkit can be present, but PyTorch cannot inspect the target card and choose an architecture automatically.

For an RTX 3090 build, the important setting is:

ENV TORCH_CUDA_ARCH_LIST="8.6"

Compute capability 8.6 is the correct target for the RTX 3090. This variable is specifically used by PyTorch CUDA extension builds. A plain nvcc or CMake CUDA project needs its equivalent compiler or CMAKE_CUDA_ARCHITECTURES setting instead.

Pytorch on Singularity

Compiling PyTorch CUDA Extensions in Docker Without a GPU