Quote:
Originally Posted by hawhill
In fact, both readers should need 0% CPU when idle and aquire up to 100% when calculating. The 22%/50% values are no valid measures - at least if you were just sitting and watching the output of "top". If you're interested in getting reliable measurements, I suggest to measure CPU time consumed. Have a look at /proc/<PID>/sched, if you want "hard numbers" :-)
(The CVM will probably have a big startup overhead, which should probably be taken into account.)
|
touché. Actually when comparing cpu usage I was merely looking at top while turning pages. It wasn't meant to be an all inclusive test. I've read that battery life is consumed when turning pages, not when idling.
cvm does have a large overhead and loaded automatically at startup.