Ganesha Bhaskara wrote: > All, > > I have been using hrt patches to minimize wakeups and increase avg > C4 residency time. With an idle system I can manage to get > 99% in > C4 with avg residency time of 120ms with 7-8 wakeup/s. This > requires 1) disabling many essential services in gnome
If gnome services are really a problem you could always try KDE. > 2) disabling DRI (I can't live without compiz ;-)) 'fraid thats life for you. I don't think there's any good reason for DRI to require a VBlank interrupt, but - for the intel driver I use at least - it's going to take a while for it to get fixed. > However, when certain essential gnome services and DRI are not > turned off, I get 16-20 wakeups/s with > 95-99% in C4 with avg > residency time of 30-50ms is. 16-20? OK that's too few to include a VBlank, you have better drivers than me :-). > Could somebody please tell me if going from the former described > configuration to the later wastes too much power ? Assuming that > one needs a usable system, what is a good range for the following > parameters and which ones are more important than the other > > 1) wakeups/s 2) avg residency time in C4 3) %age of time spent in > C4 AFAIK: The headline figure would be %age time in C4. I'm thinking the avg residency figure doesn't actually add anything to that. Transitioning to/from C4 takes some time, during which power is being drawn but no productive work done. However if you spend more time in transitions but the same time in C4, all that means is you got less time doing productive work; there might be no difference at all in terms of power usage. Hey, why are you asking us? If you've got C4, and you're mad about power consumption, aren't you on a laptop, which would have a battery meter? Powertop keeps telling me this: Suggestion: Enable the CONFIG_ACPI_BATTERY kernel configuration option. This option is required to get power estimages from PowerTOP That would give you empirical estimates of power usage for each configuration. Alan _______________________________________________ Power mailing list [email protected] http://www.bughost.org/mailman/listinfo/power
