Posted: 07 Aug 2017, 17:38
noticed something relating cpu use display, sometimes it will work as expected, lets say i use my module and it will display 50 percent cpu use,
then if i reopen the same patch/workspace it can remains blocked showing 0 percent use, then after a row of 2,3 closing reopening usine it will work back as expected, (HH2 tho, have to test on HH3)
don't know what really happens, maybe like if a core was randomly assigned and measurement relying on core id or something.
otherwise i wondered if there could be a way to debug what cpu cost is in a process, like how many time a scope needed to be achieved,
and regarding GPU, a way to monitor the impact i guess is complicated but could be interesting
then if i reopen the same patch/workspace it can remains blocked showing 0 percent use, then after a row of 2,3 closing reopening usine it will work back as expected, (HH2 tho, have to test on HH3)
don't know what really happens, maybe like if a core was randomly assigned and measurement relying on core id or something.
otherwise i wondered if there could be a way to debug what cpu cost is in a process, like how many time a scope needed to be achieved,
and regarding GPU, a way to monitor the impact i guess is complicated but could be interesting