Building Pytorch3d with CUDA12.8+pytorch2.8 on RTX 5090 GPU

[Tail-19]

RTX 5090 can only be used with torch2.8(nightly) and CUDA12.8, when will the pytorch3d support this?

I tried to use the 0.7.8 version with the install command below with torch2.6 and CUDA 12.8 but it doesn't work.

pip install --extra-index-url https://miropsota.github.io/torch_packages_builder pytorch3d==0.7.8+pt2.6.0cu126

[Tail-19]

I also tried using 0.7.8 + torch2.6 + CUDA 12.6 on 5090, but the torch is not working:
python
Python 3.9.21 (main, Dec 11 2024, 16:24:11)
[GCC 11.2.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.

import torch
torch.zeros((111000, 4, 4), device='cuda:0')
/home/hanlu/anaconda3/envs/trySNI/lib/python3.9/site-packages/torch/cuda/init.py:235: UserWarning:
NVIDIA GeForce RTX 5090 with CUDA capability sm_120 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
If you want to use the NVIDIA GeForce RTX 5090 GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/

warnings.warn(
Traceback (most recent call last):
File "", line 1, in
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

torch.cuda.is_available()
True
print(torch.version)
2.6.0+cu126

[Tail-19]

I solved the problem by simply use the mentioned comment to download pytorch3d and use torch2.8. It works.

[yeonghuiHwang]

I have a same problem

I use a RTX 5070 ti , cuda 12.8, torch 2.80 (nightly) I did a same work but, i can't solve this problem

help me : (

[yeonghuiHwang]

I have a same problem

I use a RTX 5070 ti , cuda 12.8, torch 2.80 (nightly) I did a same work but, i can't solve this problem

help me : (

thanks bro I found your comment that edited by you then I tried that.

It works well!!!!!!!!!!!!!!!!!!!!!!!

You're pretty good at developing. I appreciate for good information

[canglangzhishuiz]

I solved this issue on Windows 11 with an RTX 5070 Ti and WSL2.

First, install CUDA 12.8.
Then, install PyTorch using:

pip3 install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu128

Finally, clone the PyTorch3D source code and install it:

git clone https://github.com/facebookresearch/pytorch3d.git
cd pytorch3d
pip install -e .