Benjamin Herrenschmidt <benh@kernel.crashing.org> writes: > Did you see a significant difference in battery consumption with > different ictc values ? I had a play with something similar a couple of weeks ago. I wanted to do some proper experiments before reporting back to the list, but I haven't got around to it yet (and may not get to it until August). My plan was to produce something along the lines of <http://www.ovro.caltech.edu/~abeard/FC3_ON_8600/governors.html>, which looks at the effects of frequency scaling on a Dell laptop. In my non-scientific messing around, which looked at temperature, not battery life, it seemed to me that using ICTC resulted in a hotter chip than not using it, for a load less than 100%. My test load was just playing Ogg Vorbis files, IIRC if I used ICTC to make my iBook run with an effective execution rate of 133MHz, playing music took about 80% of CPU. Doing it this way, the CPU ran a couple of degrees hotter than if I just used powernowd to switch between 1GHz and 533MHz. cheers, John
Attachment:
pgpFtqdmLg5Cf.pgp
Description: PGP signature