You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
193.521 ms fabio
423.257 ms pyFAI.utils
435.484 ms pyFAI
1466.502 ms silx.opencl.common
1466.716 ms silx.opencl
1466.735 ms silx.opencl.common
1467.515 ms pyFAI.opencl
1570.422 ms pyFAI.load_integrators
1574.027 ms pyFAI.azimuthalIntegrator
2109.442 ms pyFAI.worker
Apparently, it is the initialization of OpenCL (which probes all devices) which takes most of the time.
The selection of the OpenCL device can be tuned with OCL_ICD_VENDORS. In this example I point to an ICD for a FPGA board that I don't have:
54.466 ms pyFAI.io.image
59.140 ms numpy.core
101.420 ms numpy
102.072 ms pyFAI.average
123.890 ms fabio.fabioimage
124.134 ms fabio.fabioformats
138.252 ms pkg_resources
138.401 ms pyFAI.resources
186.043 ms fabio
401.359 ms pyFAI.utils
414.103 ms pyFAI
558.349 ms pyFAI.worker
Condition:
ewoks
worker running onslurm
cluster; reported by Edgar.Several seconds are reported ...
The text was updated successfully, but these errors were encountered: