Distributed package doesn't have NCCL built in

vision
1

There is no error in my pytorch installation, but I am getting such an error, is it because I have only one gpu or is there a solution with a single graphics card?

2

You might be using Windows, which doesn’t support NCCL, or might have installed the CPU-only binaries. In any case, why do you want to use NCCL with a single GPU? What should the GPU communicate with?

3

Yes, I am using windows. I tried to do segmentation work with 3D point cloud data, but I encountered this error. Cuda appears but ncll gives false value, I tried reinstalling but the result did not change.

4

That’s expected as already examined since Windows does not support NCCL. However, you still didn’t answer why you want to use NCCL in the first place with a single GPU?

5

I’m trying to run a deep learning architecture that does 3d segmentation on windows without using Ubuntu, so I’m trying to use the graphic card.

6

Thank you for your interest.

7

You don’t need NCCL to run single GPU workloads, since NCCL is used for multi-GPU communication.