Vincent Favre-Nicolin

Hi @DTolm, I've updated the code so I can use batch convolution also for cuda, and also the systematic command-line test can be used for convolution. I've also clarified the...

In _principle_ it could be done - not sure how much work that would represent. However the biggest pre-requisite is a python metal library sufficiently established and maintained (like pycuda/cupy...

And it's well-tested, see on an M1 mac mini: http://ftp.esrf.fr/pub/scisoft/PyNX/pyvkfft-test/pyvkfft-test-2023-07-31-apple-m1/pyvkfft-test.html

Thanks for the report, I can confirm this behaviour also with OpenCL: ``` python import numpy as np import pyopencl as cl import pyopencl.array as cla import pyvkfft.fft from pyvkfft.base...

Hmm, I think I understand what is happening - for an axis of size 1 the stride is zero: ``` python print(x1.strides) print(x2.strides) ``` ``` (72, 24, 8) (72, 0,...

OK, this should be fixed in the current git master. The main issue was not, in fact, strides equal to zero (that's easily handled). And it is not possible to...

Thanks for the PR. I never use offsets = this corresponds to array pointing to memory allocations with an offset ? Could you add a corresponding unit test ? Nothing...

I have not tried this, so it is not implemented in python yet. @Dtolm can you clarify one point: if I read the manual correctly the zero-padding option avoids reading...

@tridao you mentioned that performing the padding required as much time as the FFT, so is it still interesting since you have to pad the array anyway ? What kind...