I've just raised a ticket on pytest-benchmark because of differences in the cpuinfo between two test runs.
It looks like hz_actual is just reporting the first value it finds in /proc/cpuinfo, i.e. a snapshot of one core. Would it be possible to take an average or min/max or something more representative? Here's what I see on my machine, the values change every time I run the command.
$ cat /proc/cpuinfo | grep "MHz"
cpu MHz : 2979.381
cpu MHz : 1401.886
cpu MHz : 4000.000
cpu MHz : 4000.000
cpu MHz : 2039.271
cpu MHz : 1797.567
cpu MHz : 1483.572
cpu MHz : 4000.000
In your bug report please include:
CPU architecture
Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz
Operating System
NAME="Ubuntu"
VERSION_ID="22.04"
VERSION="22.04.3 LTS (Jammy Jellyfish)"
Python version
3.10 (from gitlab's python:3.10 docker image)
I've just raised a ticket on
pytest-benchmark
because of differences in the cpuinfo between two test runs.It looks like
hz_actual
is just reporting the first value it finds in /proc/cpuinfo, i.e. a snapshot of one core. Would it be possible to take an average or min/max or something more representative? Here's what I see on my machine, the values change every time I run the command.In your bug report please include:
CPU architecture Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz
Operating System NAME="Ubuntu" VERSION_ID="22.04" VERSION="22.04.3 LTS (Jammy Jellyfish)"
Python version 3.10 (from gitlab's python:3.10 docker image)
Version of py-cpuinfo 9.0.0