PyTorch for Windows on Arm

PyTorch has native support for Windows on Arm . Starting with PyTorch 2.7 release, you can access Arm native builds of PyTorch for Windows available for Python 3.12.

A number of developer-ready Windows on Arm devices are available.

Windows on Arm instances are available with Microsoft Azure. For further information, see Deploy a Windows on Arm virtual machine on Microsoft Azure .

How do I install PyTorch for Windows on Arm?

Before you install PyTorch on your Windows on Arm machine, you will need to install Python version 3.12 for Windows on Arm . Select the Windows ARM64 installer.

Verify your Python installation at a Windows Command prompt or a PowerShell prompt:

    

        
        
python --version

    

The output should look like:

    

        
        Python 3.12.9

        
    

Once you have downloaded Python, you can install the PyTorch Stable release (2.7.0) on your Windows on Arm machine.

    

        
        
pip3 install torch==2.7.0 --index-url https://download.pytorch.org/whl/cpu

    

You will see that the arm64 wheel for PyTorch is installed on your machine:

    

        
        Downloading https://download.pytorch.org/whl/cpu/torch-2.7.0%2Bcpu-cp312-cp312-win_arm64.whl (107.9 MB)
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 107.9/107.9 MB 29.7 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/sympy-1.13.3-py3-none-any.whl (6.2 MB)
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 6.2/6.2 MB 47.4 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/typing_extensions-4.12.2-py3-none-any.whl (37 kB)
Downloading https://download.pytorch.org/whl/filelock-3.13.1-py3-none-any.whl (11 kB)
Downloading https://download.pytorch.org/whl/fsspec-2024.6.1-py3-none-any.whl (177 kB)
Downloading https://download.pytorch.org/whl/Jinja2-3.1.4-py3-none-any.whl (133 kB)
Downloading https://download.pytorch.org/whl/networkx-3.3-py3-none-any.whl (1.7 MB)
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.7/1.7 MB 30.6 MB/s eta 0:00:00

        
    

You can also install the nightly preview versions of PyTorch on your Windows Arm machine:

    

        
        
pip install --pre torch --index-url https://download.pytorch.org/whl/nightly/cpu

    

How can I run a PyTorch example?

To run a PyTorch example and confirm that PyTorch is working, use a text editor to save the code below to a file named pytorch_woa.py:

    

        
        
import torch
import platform

# Print PyTorch version
print("PyTorch version:", torch.__version__)

# Check if CUDA is available
if torch.cuda.is_available():
    print("CUDA is available. PyTorch can use the GPU.")
else:
    print("CUDA is not available. PyTorch will use the CPU.")

# Detect system architecture
architecture = platform.machine()
if "ARM" in architecture.upper() or "AARCH" in architecture.upper():
    print("PyTorch is running on Arm:", architecture)
else:
    print("PyTorch is not running on Arm. Detected architecture:", architecture)

# Perform a basic PyTorch operation to confirm it's working
try:
    tensor = torch.tensor([1.0, 2.0, 3.0])
    print("PyTorch is operational. Tensor created:", tensor)
except Exception as e:
    print("An error occurred while testing PyTorch:", e)

    

Run the code:

    

        
        
python pytorch_woa.py

    

Running on a Windows on Arm machine produces an output similar to:

    

        
        PyTorch version: 2.7.0+cpu
CUDA is not available. PyTorch will use the CPU.
PyTorch is running on Arm: ARM64
PyTorch is operational. Tensor created: tensor([1., 2., 3.])

        
    

You are now ready to use PyTorch on your Windows on Arm device.