The following fails:
Python 3.7.2 (default, Jan 10 2019, 23:51:51)
Type 'copyright', 'credits' or 'license' for more information
IPython 7.1.1 -- An enhanced Interactive Python. Type '?' for help.
In [1]: import numpy as np
In [2]: import pycuda
In [3]: import pycuda.driver as drv
---------------------------------------------------------------------------
ModuleNotFoundError Traceback (most recent call last)
ModuleNotFoundError: No module named 'numpy.core._multiarray_umath'
terminate called after throwing an instance of 'std::runtime_error'
what(): numpy failed to initialize
zsh: abort (core dumped) ipython
numpy version 1.15.4
python version 3.7.2
Thanks,
Edwin
Hi,
We use pycuda to access NVENC and our application (xpra) if very
sensitive to latency, unfortunately it seems that pycuda will hold the
GIL during driver.init() and driver.make_context() and those calls can
take hundreds of milliseconds to complete.
Is there any reason why this is the case?
Would you consider a change to pycuda to release the GIL?
Thanks,
Antoine