Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Detect GPU and `import cupy as np`

See original GitHub issue

I could not find an example how to detect if GPU is present and swap numpy for cupy during import. Is that possible?

from cupy import gpu
if gpu:
  import cupy as np
else:
  import numpy as np

Issue Analytics

State:
Created 2 years ago
Comments:9 (4 by maintainers)

Top GitHub Comments

1reaction

kmaehashicommented, Aug 26, 2021

I am afraid that doesn’t answer the question, and it is still the catch everything antipattern.

I agree that catch-all is an antipattern but currently except Exception is the safest way. Please see https://github.com/cupy/cupy/issues/2148 for the context.

If there a plan to move non-specific functions into cupy.gpu namespace?

We are using HIP which is a CUDA-compatible layer for AMD/ROCm, so we currently assume everything to be common between two.

1reaction

kmaehashicommented, Aug 23, 2021

cupy.cuda.is_available() returns True if 1 or more devices are available.

import numpy as np
try:
    import cupy
    if cupy.cuda.is_available():
        np = cupy
except:
    pass

Top Results From Across the Web

Basics of CuPy — CuPy 11.4.0 documentation

CuPy is a GPU array backend that implements a subset of NumPy interface. In the following code, ... import numpy as np >>>...

How to run python on GPU with CuPy? - Stack Overflow

However, when I run nvidia-smi , no GPU processes are found. nvidia-smi output. Here's the code: import numpy as np import cupy as...

Using your GPU with CuPy - The Carpentries Incubator

CuPy is a GPU array library that implements a subset of the NumPy and SciPy ... import numpy as np # Construct an...

Here's How to Use CuPy to Make Numpy Over 10X Faster

CuPy is a library that implements Numpy arrays on Nvidia GPUs by leveraging the CUDA GPU library. With that implementation, superior ...

How to use the cupy.cuda.Device function in cupy - Snyk

import cupy as cp import numpy as np import torch import torch.nn.functional as F from mxtorch.vision.bbox_tools import loc2bbox from torch import nn from ......