An Argonne National Laboratory researcher has written a parsing tool to summarize MKL_VERBOSE output.
The tool can be very useful for customers who need summary of many MKL calls and their statistics.
The link to the GitHub is https://github.com/TApplencourt/mkl-verbose-toolkit
pip install -r requirements.txt
conda install --file requirements.txt
mkl_parse.py [-h] [-n #] [filename]
Generate a summary for "MKL_verbosed" log files. $MKLROOT will be used to match MKL arguments name with respective values.
Positional arguments: filename If filename is not provided, std.in will be used
Optional arguments: -h, --help show this help message and exit -n #, --num_routines # The number of routines to print out
For more complete information about compiler optimizations, see our Optimization Notice.