Hi, error with last step setting up: pip install pointnet2_ops_lib/.

[yiliu-coding]

  C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.3\include\crt/host_config.h(160): fatal error C1189: #error:  -- unsupported Microsoft Visual Studio version! Only the versions between 2017 and 2019 (inclusive) are supported! The nvcc flag '-allow-unsupported-compiler' can be used to override this version check; however, using an unsupported host compiler may cause compilation failure or incorrect run time execution. Use at your own risk.
  nvcc warning : The 'compute_35', 'compute_37', 'compute_50', 'sm_35', 'sm_37' and 'sm_50' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
  ball_query_gpu.cu
  error: command 'C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v11.3\\bin\\nvcc.exe' failed with exit code 2
  [end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.
error: legacy-install-failure

× Encountered error while trying to install package.
╰─> pointnet2-ops

note: This is an issue with the package mentioned above, not pip.
hint: See above for output from the failure.

[ZrrSkywalker]

Hi @yiliu-coding , seems that it's because of the cuda version or windows system. Have you tried on Linux?
Or you can refer to this issue for an alternative of pointnet2-ops.

[Cipher-zzz]

@ZrrSkywalker @yiliu-coding I got the same issue in Linux with other versions of CUDA and Pytorch.

Here is how I solved it:

Install CUDA 11.4 (11. x should be fine, but I haven't tested all of them);

Install this typical Pytorch version 1.7.1+cu110:

pip install torch==1.7.1+cu110 torchvision==0.8.2+cu110 torchaudio==0.7.2 -f https://download.pytorch.org/whl/torch_stable.html

Then the pointnet2-ops installation works fine in both Windows 10 and Ubuntu 18.04.

[ZrrSkywalker]

@Cipher-zzz Thanks very much for your help :).

[jiaxiaovii]

Hi @ZrrSkywalker @yiliu-coding I got the same issue in Linux with other versions of CUDA and Pytorch.

I encountered a similar issue with the configuration you suggested on my Ubuntu system. Despite following the steps and installing CUDA 11.4 along with PyTorch 1.7.1+cu110, I still faced problems during the installation of pointnet2-ops. Additionally, I attempted to use other versions of CUDA including 11.1, 11.3, 11.7, and 11.8, but encountered the same issue. It seems like there might be a compatibility issue or another underlying problem.