How to beginner · 3 min read

How to check if CUDA is available in PyTorch

Q: How to check if CUDA is available in PyTorch

Use torch.cuda.is_available() to check if CUDA is available in PyTorch. It returns True if a compatible GPU and CUDA driver are detected, otherwise False.

Quick answer

Use torch.cuda.is_available() to check if CUDA is available in PyTorch. It returns True if a compatible GPU and CUDA driver are detected, otherwise False.

PREREQUISITES

Python 3.8+
pip install torch

Setup

Install PyTorch with CUDA support by following the official instructions at PyTorch official site. For most users, installing the latest stable version with CUDA support is recommended.

Example install command for CUDA 11.8:

bash

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

Step by step

Run the following Python code to check if CUDA is available on your system. This code imports torch and prints the CUDA availability status.

python

import torch

if torch.cuda.is_available():
    print("CUDA is available. You can use GPU acceleration.")
else:
    print("CUDA is not available. Using CPU.")

output

CUDA is available. You can use GPU acceleration.

Common variations

You can also check the number of available CUDA devices and their names for more detailed info:

python

import torch

def print_cuda_info():
    if torch.cuda.is_available():
        print(f"Number of CUDA devices: {torch.cuda.device_count()}")
        for i in range(torch.cuda.device_count()):
            print(f"Device {i}: {torch.cuda.get_device_name(i)}")
    else:
        print("CUDA is not available.")

print_cuda_info()

output

Number of CUDA devices: 1
Device 0: NVIDIA GeForce RTX 3080

Troubleshooting

If torch.cuda.is_available() returns False but you expect CUDA to be available, check the following:

Ensure your GPU drivers and CUDA toolkit are properly installed and compatible with your PyTorch version.
Verify that you installed the PyTorch version with CUDA support, not the CPU-only version.
Restart your machine after installing drivers or CUDA toolkit.
Check environment variables like CUDA_VISIBLE_DEVICES that might disable GPU visibility.

✅

Key Takeaways

Use torch.cuda.is_available() to quickly verify CUDA support in PyTorch.
Check device count and names with torch.cuda.device_count() and torch.cuda.get_device_name() for detailed GPU info.
Make sure to install the PyTorch version with CUDA support matching your CUDA toolkit and GPU drivers.
If CUDA is not detected, verify driver installation and environment variables before troubleshooting further.

Verified 2026-04

Verify ↗