# windows

**URL:** https://discuss.pytorch.org/c/windows/26.md

[Latest](https://discuss.pytorch.org/latest.md) · [Categories](https://discuss.pytorch.org/categories.md)

---

## [About the windows category](https://discuss.pytorch.org/t/about-the-windows-category/89244)

<div class="topic-metadata">

**Author:** [@jspisak](https://discuss.pytorch.org/u/jspisak)\
**Replies:** 0

</div>

This category is focused on PyTorch on Windows related issues.

---

## [PyTorch cannot detect RTX 5080 (requires CUDA 12.8) – Nightly build fails to install correctly on Windows](https://discuss.pytorch.org/t/pytorch-cannot-detect-rtx-5080-requires-cuda-12-8-nightly-build-fails-to-install-correctly-on-windows/223957)

<div class="topic-metadata">

**Author:** [@tai353](https://discuss.pytorch.org/u/tai353)\
**Replies:** 3\
**Last updated:** [July 14, 2026, 4:35pm UTC](https://discuss.pytorch.org/t/pytorch-cannot-detect-rtx-5080-requires-cuda-12-8-nightly-build-fails-to-install-correctly-on-windows/223957 "2026-07-14T16:35:02Z")

</div>

Hello, I need help with installing PyTorch for my new NVIDIA RTX 5080 on Windows. According to the PyTorch contributor reply on GitHub, the oldest CUDA version that supports RTX 5080 is CUDA 12.8. They also said that a…

---

## [How Do I use Pytorch with RTX 5060 Ti](https://discuss.pytorch.org/t/how-do-i-use-pytorch-with-rtx-5060-ti/220926)

<div class="topic-metadata">

**Author:** [@Jeff\_H](https://discuss.pytorch.org/u/Jeff_H)\
**Replies:** 19\
**Last updated:** [June 1, 2026, 7:19pm UTC](https://discuss.pytorch.org/t/how-do-i-use-pytorch-with-rtx-5060-ti/220926 "2026-06-01T19:19:43Z")

</div>

Hello I’m new here. Can someone help me with this? I can’t find a compatible pytorch for my RTX 5060 Ti. Here’s my system and it’s issues. I’ve tried the nightly version and it still won’t work. "Python 3.10.11 Checkin…

---

## [Unable to install pytorch 2.10 using conda 26.3.2 on Windows 11](https://discuss.pytorch.org/t/unable-to-install-pytorch-2-10-using-conda-26-3-2-on-windows-11/224920)

<div class="topic-metadata">

**Author:** [@shaynegw](https://discuss.pytorch.org/u/shaynegw)\
**Replies:** 3\
**Last updated:** [May 11, 2026, 2:07pm UTC](https://discuss.pytorch.org/t/unable-to-install-pytorch-2-10-using-conda-26-3-2-on-windows-11/224920 "2026-05-11T14:07:35Z")

</div>

Hi Has anyone had issues installing pytorch using conda? We have an on-premises Anacoda repo (air gapped /w no Internet) and I am unable to install pytorch using ‘conda install pytorch’ in a clean environment. conda cr…

---

## [PyTorch not recognising CUDA device](https://discuss.pytorch.org/t/pytorch-not-recognising-cuda-device/224742)

<div class="topic-metadata">

**Author:** [@g.j.paz](https://discuss.pytorch.org/u/g.j.paz)\
**Replies:** 2\
**Last updated:** [March 27, 2026, 3:17pm UTC](https://discuss.pytorch.org/t/pytorch-not-recognising-cuda-device/224742 "2026-03-27T15:17:41Z")

</div>

I’ve installed anaconda 26.1.1 on windows 10, and am trying to get PyTorch to work on my workstation with a Quadro P5000. I have installed cudatoolkit, cudnn and pytorch, but my system is not recognising the cuda device…

---

## [RuntimeError: Couldn't install torch.](https://discuss.pytorch.org/t/runtimeerror-couldnt-install-torch/224630)

<div class="topic-metadata">

**Author:** [@gubatsonpulta](https://discuss.pytorch.org/u/gubatsonpulta)\
**Replies:** 2\
**Last updated:** [March 3, 2026, 12:03am UTC](https://discuss.pytorch.org/t/runtimeerror-couldnt-install-torch/224630 "2026-03-03T00:03:41Z")

</div>

I’ve been trying following instructions here, https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-AMD-GPUs for amd gpu on windows but keep running into the same issue. venv "C:\\Users\\rob\\Des…

---

## [Pytorch dev 130](https://discuss.pytorch.org/t/pytorch-dev-130/224539)

<div class="topic-metadata">

**Author:** [@Hayri](https://discuss.pytorch.org/u/Hayri)\
**Replies:** 2\
**Last updated:** [February 13, 2026, 3:28pm UTC](https://discuss.pytorch.org/t/pytorch-dev-130/224539 "2026-02-13T15:28:39Z")

</div>

When I try to train with the new dev pytorch, it stops without giving an error message. The old version, i.e. 129, said something about gloo but it seems to have been fixed in the latest release, but it can’t train.

---

## [Uninstall PyTorch completely to install older version](https://discuss.pytorch.org/t/uninstall-pytorch-completely-to-install-older-version/223551)

<div class="topic-metadata">

**Author:** [@JustWriteSomething](https://discuss.pytorch.org/u/JustWriteSomething)\
**Replies:** 11\
**Last updated:** [January 21, 2026, 2:43pm UTC](https://discuss.pytorch.org/t/uninstall-pytorch-completely-to-install-older-version/223551 "2026-01-21T14:43:04Z")

</div>

I have the kernel error in ComyUI that suggests my current PyTorch install (v12.8 on Windows 10) is incompatible with my GeForce GTX 960 (CUDA capability 5.2) and CUDA 12.6. When I have tried to uninstall PyTorch and in…

---

## [Impossible to install for CUDA 12.8](https://discuss.pytorch.org/t/impossible-to-install-for-cuda-12-8/224319)

<div class="topic-metadata">

**Author:** [@Bloody\_KGB](https://discuss.pytorch.org/u/Bloody_KGB)\
**Replies:** 5\
**Last updated:** [January 12, 2026, 10:09pm UTC](https://discuss.pytorch.org/t/impossible-to-install-for-cuda-12-8/224319 "2026-01-12T22:09:23Z")

</div>

I have python 3.13.9 and CUDA 12.8. Command install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128 returns ERROR: Could not find a version that satisfies the requirement torch (from versi…

---

## [Torch 2.9.1 cu129 cp312?](https://discuss.pytorch.org/t/torch-2-9-1-cu129-cp312/224340)

<div class="topic-metadata">

**Author:** [@langfod](https://discuss.pytorch.org/u/langfod)\
**Replies:** 1\
**Last updated:** [January 12, 2026, 9:35pm UTC](https://discuss.pytorch.org/t/torch-2-9-1-cu129-cp312/224340 "2026-01-12T21:35:45Z")

</div>

Just out of curiosity, why is there no Windows wheel for torch-2.9.1+cu129-cp312-cp312-win\_amd64.whl ? To update torch requires downgrading to cu128?

---

## [UserWarning: CUDA initialization: The NVIDIA driver on your system is too old (found version 10010)](https://discuss.pytorch.org/t/userwarning-cuda-initialization-the-nvidia-driver-on-your-system-is-too-old-found-version-10010/141547)

<div class="topic-metadata">

**Author:** [@dan\_tyvan](https://discuss.pytorch.org/u/dan_tyvan)\
**Replies:** 19\
**Last updated:** [January 7, 2026, 3:58pm UTC](https://discuss.pytorch.org/t/userwarning-cuda-initialization-the-nvidia-driver-on-your-system-is-too-old-found-version-10010/141547 "2026-01-07T15:58:39Z")

</div>

Hi Everyone, I’m having trouble with this error. The error message is not quite clear to me, in one part it says NVIDIA driver, and in the other part it says CUDA driver, and each time I am not sure what they are referr…

---

## [Python 3 pytorch](https://discuss.pytorch.org/t/python-3-pytorch/224072)

<div class="topic-metadata">

**Author:** [@Hayri](https://discuss.pytorch.org/u/Hayri)\
**Replies:** 3\
**Last updated:** [December 9, 2025, 3:41am UTC](https://discuss.pytorch.org/t/python-3-pytorch/224072 "2025-12-09T03:41:17Z")

</div>

How can I update python without installing it? Can I use a zip file for that?

---

## [Pytorch system memmory](https://discuss.pytorch.org/t/pytorch-system-memmory/224012)

<div class="topic-metadata">

**Author:** [@Hayri](https://discuss.pytorch.org/u/Hayri)\
**Replies:** 4\
**Last updated:** [November 27, 2025, 2:49pm UTC](https://discuss.pytorch.org/t/pytorch-system-memmory/224012 "2025-11-27T14:49:15Z")

</div>

why can’t pytorch 2.7.1 use system memory?

---

## [Pytorch 128 and 129](https://discuss.pytorch.org/t/pytorch-128-and-129/223953)

<div class="topic-metadata">

**Author:** [@Hayri](https://discuss.pytorch.org/u/Hayri)\
**Replies:** 5\
**Last updated:** [November 15, 2025, 10:51pm UTC](https://discuss.pytorch.org/t/pytorch-128-and-129/223953 "2025-11-15T22:51:30Z")

</div>

I am using google translater: I have a problem with the next-newest pytorch 128 alts¨2.8.0 When I train in RVC I see that white glow is not supported device. Also the old pytorch 118 alts¨2.0 could allocate my ram autom…

---

## [TorchCodec wheels](https://discuss.pytorch.org/t/torchcodec-wheels/218578)

<div class="topic-metadata">

**Author:** [@richbill](https://discuss.pytorch.org/u/richbill)\
**Replies:** 2\
**Last updated:** [October 23, 2025, 11:22am UTC](https://discuss.pytorch.org/t/torchcodec-wheels/218578 "2025-10-23T11:22:55Z")

</div>

I want to install torchcodec 0.2.1 on windows. I checked download.pytorch.org/whl/nightly/torchcodec/ Are all the versions non-windows? Is there one I should use or could someone make one?

---

## [CUDA profiler not working on Windows](https://discuss.pytorch.org/t/cuda-profiler-not-working-on-windows/223644)

<div class="topic-metadata">

**Author:** [@MLML](https://discuss.pytorch.org/u/MLML)\
**Replies:** 0\
**Last updated:** [October 18, 2025, 5:54am UTC](https://discuss.pytorch.org/t/cuda-profiler-not-working-on-windows/223644 "2025-10-18T05:54:05Z")

</div>

Hi, I am new to pytorch and trying to set up a pytorch + GPU env on windows but currently stuck in a state where CUDA works (torch.cuda.is\_available() = true) but torch.profiler does not work due to the following error: …

---

## [PyTorch scripts not on PATH](https://discuss.pytorch.org/t/pytorch-scripts-not-on-path/222873)

<div class="topic-metadata">

**Author:** [@lfd.arch](https://discuss.pytorch.org/u/lfd.arch)\
**Replies:** 1\
**Last updated:** [September 30, 2025, 3:08pm UTC](https://discuss.pytorch.org/t/pytorch-scripts-not-on-path/222873 "2025-09-30T15:08:05Z")

</div>

I am trying to install PyTorch for ComfyUI. I just uninstalled it and reinstalled it because I thought it didn’t work. Now, so far it works because I checked with the torch commands in the “get started” page and it conne…

---

## [Windows linker error](https://discuss.pytorch.org/t/windows-linker-error/223316)

<div class="topic-metadata">

**Author:** [@annoyed](https://discuss.pytorch.org/u/annoyed)\
**Replies:** 0\
**Last updated:** [September 25, 2025, 1:06pm UTC](https://discuss.pytorch.org/t/windows-linker-error/223316 "2025-09-25T13:06:58Z")

</div>

Im trying to building torch locally on Windows and I can’t get past this linker error: FAILED: bin/torch\_cpu.dll lib/torch\_cpu.lib cmd.exe /C “cd . && “C:\\Program Files\\CMake\\bin\\cmake.exe” -E vs\_link\_dll --intdir=caff…

---

## [RTX 5070 Ti (Blackwell) + PyTorch Nightly + Triton: Still Getting "sm\_120 is not defined for option 'gpu-name'" Error](https://discuss.pytorch.org/t/rtx-5070-ti-blackwell-pytorch-nightly-triton-still-getting-sm-120-is-not-defined-for-option-gpu-name-error/220460)

<div class="topic-metadata">

**Author:** [@NexTGenioN](https://discuss.pytorch.org/u/NexTGenioN)\
**Replies:** 5\
**Last updated:** [September 24, 2025, 1:24pm UTC](https://discuss.pytorch.org/t/rtx-5070-ti-blackwell-pytorch-nightly-triton-still-getting-sm-120-is-not-defined-for-option-gpu-name-error/220460 "2025-09-24T13:24:12Z")

</div>

Hello everyone, I’m trying to get ComfyUI working on a new RTX 5070 Ti (Blackwell) GPU under Windows 11. I’ve already upgraded to the latest PyTorch Nightly with CUDA 12.8 support (which should include Blackwell suppor…

---

## [\[Question\]CUDA error: no kernel image is available for execution on the device](https://discuss.pytorch.org/t/question-cuda-error-no-kernel-image-is-available-for-execution-on-the-device/223122)

<div class="topic-metadata">

**Author:** [@fiona-fifi](https://discuss.pytorch.org/u/fiona-fifi)\
**Replies:** 1\
**Last updated:** [September 16, 2025, 3:40pm UTC](https://discuss.pytorch.org/t/question-cuda-error-no-kernel-image-is-available-for-execution-on-the-device/223122 "2025-09-16T15:40:39Z")

</div>

Hi, I am trying to install and use PyTorch on my machine, but something went wrong: AcceleratorError: CUDA error: no kernel image is available for execution on the device CUDA kernel errors might be asynchronously repor…

---

## [How am i supposed to get old versions of torch on windows x64 arhitecture?](https://discuss.pytorch.org/t/how-am-i-supposed-to-get-old-versions-of-torch-on-windows-x64-arhitecture/222813)

<div class="topic-metadata">

**Author:** [@Alex\_Eugen\_Botez](https://discuss.pytorch.org/u/Alex_Eugen_Botez)\
**Replies:** 2\
**Last updated:** [September 5, 2025, 9:36am UTC](https://discuss.pytorch.org/t/how-am-i-supposed-to-get-old-versions-of-torch-on-windows-x64-arhitecture/222813 "2025-09-05T09:36:12Z")

</div>

All posts lead me to https://download.pytorch.org/whl/cu80/torch\_stable.html, on which i can’t seem to find any wheels for windows x64. Or how can i get a version of pytorch running for my 8.6 compute GPU?

---

## [CUDA toolkit 13.0 uninstalled but torch.utils.cpp\_extension gives CUDA\_HOME of v13](https://discuss.pytorch.org/t/cuda-toolkit-13-0-uninstalled-but-torch-utils-cpp-extension-gives-cuda-home-of-v13/222694)

<div class="topic-metadata">

**Author:** [@alicemabille](https://discuss.pytorch.org/u/alicemabille)\
**Replies:** 1\
**Last updated:** [August 27, 2025, 4:21pm UTC](https://discuss.pytorch.org/t/cuda-toolkit-13-0-uninstalled-but-torch-utils-cpp-extension-gives-cuda-home-of-v13/222694 "2025-08-27T16:21:18Z")

</div>

I installed CUDA toolkit v12.9, then uninstalled CUDA toolkit v13.0, deleted C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v13.0, checked that no environment variables had this version, nvcc –-version gives nvcc: N…

---

## [I successfully built PyTorch from source on Windows 11 using CUDA 12.9 for the RTX 5070. Still working through a small runtime issue, but the full build process is documented here on GitHub: \[https://github.com/DMorford/TorchinDesiree\]](https://discuss.pytorch.org/t/i-successfully-built-pytorch-from-source-on-windows-11-using-cuda-12-9-for-the-rtx-5070-still-working-through-a-small-runtime-issue-but-the-full-build-process-is-documented-here-on-github-https-github-com-dmorford-torchindesiree/221817)

<div class="topic-metadata">

**Author:** [@EasydoesitDave](https://discuss.pytorch.org/u/EasydoesitDave)\
**Replies:** 1\
**Last updated:** [August 13, 2025, 6:24pm UTC](https://discuss.pytorch.org/t/i-successfully-built-pytorch-from-source-on-windows-11-using-cuda-12-9-for-the-rtx-5070-still-working-through-a-small-runtime-issue-but-the-full-build-process-is-documented-here-on-github-https-github-com-dmorford-torchindesiree/221817 "2025-08-13T18:24:47Z")

</div>

========================================= PYTORCH RTX 5070 BUILD SESSION SUMMARY Date: July 24, 2025 Machine: Desiree (i9-14900F, 32GB RAM, RTX 5070 12GB) CUDA Version: 12.9 Build Status: COMPILATION SUCCESSFUL …

---

## [NVIDIA GeForce RTX 5090 with CUDA capability sm\_120 is not compatible with the current PyTorch installation](https://discuss.pytorch.org/t/nvidia-geforce-rtx-5090-with-cuda-capability-sm-120-is-not-compatible-with-the-current-pytorch-installation/216518)

<div class="topic-metadata">

**Author:** [@viyriy](https://discuss.pytorch.org/u/viyriy)\
**Replies:** 21\
**Last updated:** [July 29, 2025, 4:04pm UTC](https://discuss.pytorch.org/t/nvidia-geforce-rtx-5090-with-cuda-capability-sm-120-is-not-compatible-with-the-current-pytorch-installation/216518 "2025-07-29T16:04:29Z")

</div>

I keep getting this error: torch\\cuda\_init\_.py:230: UserWarning: NVIDIA GeForce RTX 5090 with CUDA capability sm\_120 is not compatible with the current PyTorch installation. The current PyTorch install supports CUDA c…

---

## [optim.Adam error](https://discuss.pytorch.org/t/optim-adam-error/221784)

<div class="topic-metadata">

**Author:** [@mohmead](https://discuss.pytorch.org/u/mohmead)\
**Replies:** 3\
**Last updated:** [July 24, 2025, 10:26pm UTC](https://discuss.pytorch.org/t/optim-adam-error/221784 "2025-07-24T22:26:51Z")

</div>

i get File :1 def create\_fn(dataclass\_type\_nx-loopback, dataclass\_HAS\_DEFAULT\_FACTORY, dataclass\_builtins\_object, dataclass\_return\_type): ^ SyntaxError: invalid syntax when use any optimizer such that optimizer = …

---

## [Pytorch Profiler not profiling GPU on Windows](https://discuss.pytorch.org/t/pytorch-profiler-not-profiling-gpu-on-windows/146685)

<div class="topic-metadata">

**Author:** [@ChowderII](https://discuss.pytorch.org/u/ChowderII)\
**Replies:** 19\
**Last updated:** [July 21, 2025, 12:32am UTC](https://discuss.pytorch.org/t/pytorch-profiler-not-profiling-gpu-on-windows/146685 "2025-07-21T00:32:08Z")

</div>

Hello everyone, I’m new here, hopefully I write this in the correct way. I’ve recently gotten to use PyTorch’s profiler but I can’t seem to see any activity on my GPU as far as the profiler is concerned. Currently I’m …

---

## [How to use torch with CUDA 12.9](https://discuss.pytorch.org/t/how-to-use-torch-with-cuda-12-9/220827)

<div class="topic-metadata">

**Author:** [@Ahmed\_Mhmouad](https://discuss.pytorch.org/u/Ahmed_Mhmouad)\
**Replies:** 4\
**Last updated:** [July 18, 2025, 10:41pm UTC](https://discuss.pytorch.org/t/how-to-use-torch-with-cuda-12-9/220827 "2025-07-18T22:41:59Z")

</div>

I’m running with the following environment: Windows 10 python 3.12.0 CUDA Version: 12.9 (according to \`nvidia-smi\`) torch: 2.7.1 I did not find 12.9 at installation settings so i choose the nearest version 12.8 I’m r…

---

## [Windows & WSL2: zeroed CUDA tensors in spawned processes](https://discuss.pytorch.org/t/windows-wsl2-zeroed-cuda-tensors-in-spawned-processes/220788)

<div class="topic-metadata">

**Author:** [@taras-janea](https://discuss.pytorch.org/u/taras-janea)\
**Replies:** 0\
**Last updated:** [June 13, 2025, 9:34am UTC](https://discuss.pytorch.org/t/windows-wsl2-zeroed-cuda-tensors-in-spawned-processes/220788 "2025-06-13T09:34:02Z")

</div>

Dear PyTorch Team, Please help to find the root cause of the issue described below. Issue description Executing the following script using the PyTorch 2.7, CUDA 12.6: import torch import torch.multiprocessing as mp …

---

## [Torch errors on Intel GPU](https://discuss.pytorch.org/t/torch-errors-on-intel-gpu/220386)

<div class="topic-metadata">

**Author:** [@Karakurt](https://discuss.pytorch.org/u/Karakurt)\
**Replies:** 1\
**Last updated:** [May 28, 2025, 3:41am UTC](https://discuss.pytorch.org/t/torch-errors-on-intel-gpu/220386 "2025-05-28T03:41:01Z")

</div>

I am trying to run some things on my laptop with Intel GPU using Ubuntu 24.04 under WSL2 in Windows 11 24H2. As I found at the Pytorch website, it should be painless enough by just changing device to ‘xpu’. Of course I’v…

---

## [Windows Debug Build GIL Error When Importing Torch](https://discuss.pytorch.org/t/windows-debug-build-gil-error-when-importing-torch/216021)

<div class="topic-metadata">

**Author:** [@ikalinic](https://discuss.pytorch.org/u/ikalinic)\
**Replies:** 1\
**Last updated:** [May 25, 2025, 5:08pm UTC](https://discuss.pytorch.org/t/windows-debug-build-gil-error-when-importing-torch/216021 "2025-05-25T17:08:10Z")

</div>

Hi everyone. Does anyone know the correct way of building PyTorch in Debug mode on Windows? I’ve been able to build using set DEBUG=1 and running python setup.py develop with also setting the LIB env variable to the Pyth…

[Next page](https://discuss.pytorch.org/c/windows/26.md?page=1)
