I get the following error when trying to import CuPy:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python2.7/dist-packages/cupy/__init__.py", line 20, in <module>
    six.reraise(ImportError, ImportError(msg), exc_info[2])
  File "/usr/local/lib/python2.7/dist-packages/cupy/__init__.py", line 9, in <module>
    from cupy import core  # NOQA
  File "/usr/local/lib/python2.7/dist-packages/cupy/core/__init__.py", line 1, in <module>
    from cupy.core import core  # NOQA
  File "cupy/core/core.pyx", line 1, in init cupy.core.core (cupy/core/core.cpp:103472)
  File "/usr/local/lib/python2.7/dist-packages/cupy/cuda/__init__.py", line 3, in <module>
    from cupy.cuda import compiler  # NOQA
  File "/usr/local/lib/python2.7/dist-packages/cupy/cuda/compiler.py", line 12, in <module>
    from cupy.cuda import function
  File "cupy/cuda/function.pyx", line 1, in init cupy.cuda.function (cupy/cuda/function.cpp:6346)
ImportError: CuPy is not correctly installed. Please check your environment, uninstall CuPy and reinstall it with `pip install cupy --no-cache-dir -vvvv`.

original error: libnvidia-fatbinaryloader.so.375.82: cannot open shared object file: No such file or directory

I have tried to reinstall CuPy with the suggestion. Also, I tried to copy the libnvidia-fatbinaryloader.so.375.82 file from /usr/lib/nvidia-375 to /usr/local/cuda-8.0/lib64/ but does not work.

In fact, when I copy that file to /usr/local/cuda-8.0/lib64/ I get this error (when trying to create an array):

>>> import cupy as cp

>>> x_gpu = cp.array([1, 2, 3])

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python2.7/dist-packages/cupy/creation/from_data.py", line 26, in array
    return core.array(obj, dtype, copy, ndmin)
  File "cupy/core/core.pyx", line 1822, in cupy.core.core.array (cupy/core/core.cpp:59154)
  File "cupy/core/core.pyx", line 1848, in cupy.core.core.array (cupy/core/core.cpp:58918)
  File "cupy/core/core.pyx", line 92, in cupy.core.core.ndarray.__init__ (cupy/core/core.cpp:6509)
  File "cupy/cuda/memory.pyx", line 283, in cupy.cuda.memory.alloc (cupy/cuda/memory.cpp:5911)
  File "cupy/cuda/memory.pyx", line 263, in cupy.cuda.memory._malloc (cupy/cuda/memory.cpp:5853)
  File "cupy/cuda/memory.pyx", line 264, in cupy.cuda.memory._malloc (cupy/cuda/memory.cpp:5739)
  File "cupy/cuda/memory.pyx", line 34, in cupy.cuda.memory.Memory.__init__ (cupy/cuda/memory.cpp:1789)
  File "cupy/cuda/device.pyx", line 67, in cupy.cuda.device.Device.__init__ (cupy/cuda/device.cpp:1789)
  File "cupy/cuda/runtime.pyx", line 158, in cupy.cuda.runtime.getDevice (cupy/cuda/runtime.cpp:2563)
  File "cupy/cuda/runtime.pyx", line 130, in cupy.cuda.runtime.check_status (cupy/cuda/runtime.cpp:2242)
cupy.cuda.runtime.CUDARuntimeError: cudaErrorUnknown: unknown error

My CUDA version: 8.0

Ubuntu 16.04

Python: 2.7

Could somebody help? I want to avoid reinstalling CUDA if it's possible. If you need further information to solve it, do not hesitate to ask. Thank you!

  • 2
    How exactly did you install CuPy? What's the ouput of the diagnostic command pip install cupy --no-cache-dir -vvvv mentioned in the error message? – David Foerster Oct 17 '17 at 7:14

I reinstalled the NVIDIA drivers and it worked out..

  • 3
    This answer would be more useful to those with the same issue without your experience if you were to go into detail regarding how you reinstalled the NVIDIA drivers and how it worked out. Please review askubuntu.com/help/how-to-answer – Elder Geek Oct 6 '17 at 21:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.