I'm running Ubuntu 13.10 on a MSI gs70 Stealth laptop. When I run powertop 2.4, I see:

The battery reports a discharge rate of 38.1 W The estimated remaining time is 0 hours, 5 minutes

Summary: 1524,0 wakeups/second,  56,0 GPU ops/seconds, 0,0 VFS ops/sec and 28,3% CPU use

Power est.              Usage       Events/s    Category       Description
  6.97 W     96,7%                      Device         Display backlight
  1.19 W     18,9 ms/s     296,8        Interrupt      [6] tasklet(softirq)   586 mW      1,2 ms/s     146,8        kWork         od_dbs_timer   525 mW      7,0 ms/s     130,8        Process        smbd -F   454 mW      3,1 ms/s     113,4        Timer          hrtimer_wakeup   450 mW      3,4 ms/s     113,5        Process        /usr/bin/pulseaudio --start --log-target=syslog   445 mW     10,5 ms/s 100,6        Process        /opt/google/chrome/chrome
--no-startup-window   398 mW     20,4 ms/s      91,5        Process        compiz   369 mW      1,1 ms/s      92,4        Process        [rcu_sched]   323 mW     79,0 ms/s      72,7        Process        /opt/google/chrome/chrome --type=renderer --lang=en-US
--force-fieldtrials=AutocompleteDynamicTrial_2/DefaultControl_R2_Stable/ 301 mW     35,1 ms/s      67,6        Process        /usr/bin/tilda   230 mW      1,8 ms/s      57,5        Timer          tick_sched_timer  163 mW      3,9 ms/s      40,4        Interrupt      [46] i915   158 mW     10,7 ms/s      38,5        Process        /usr/bin/ibus-daemon
--daemonize --xim
 89.6 mW     21,9 ms/s      20,1        Process        /usr/bin/X -core :0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch
 86.9 mW      2,0 ms/s      21,6        Interrupt      PS/2 Touchpad / Keyboard / Mouse
 73.7 mW      3,0 ms/s      18,1        Process        /usr/lib/ibus/ibus-engine-simple
 69.0 mW    169,3 µs/s      17,3        kWork          flush_to_ldisc
 67.2 mW    149,9 µs/s      16,8        kWork          intel_unpin_work_fn
 47.2 mW     11,5 ms/s      10,6        Process        /usr/lib/unity/unity-panel-service
 43.3 mW    321,8 µs/s      10,8        Process        java -jar Downloads/tomighty-0.7.1.jar
 42.3 mW      3,7 ms/s      10,2        Process        /opt/google/chrome/chrome --type=renderer --lang=en-US
 30.7 mW      6,9 ms/s       7,0        Process        powertop
 30.7 mW      1,0 ms/s       7,6        Process        zeitgeist-datahub
 29.8 mW      1,4 ms/s       7,3        Process        /usr/lib/x86_64-linux-gnu/indicator-application-service
 28.0 mW     36,6 µs/s       7,0        kWork          console_callback
 26.8 mW      3,4 ms/s       6,4        Process        indicator-multiload
 25.0 mW    126,9 µs/s       5,6        Process        syndaemon -i 1.0 -t -K -R
 23.4 mW     87,2 µs/s       5,8        kWork          ieee80211_iface_work
 20.4 mW      6,7 ms/s       4,4        Process        /opt/google/chrome/chrome --type=renderer --lang=en-US
 19.9 mW    236,9 µs/s       5,0        Process        [rcuos/7]
 16.6 mW    648,9 µs/s       4,1        Process        /usr/lib/x86_64-linux-gnu/bamf/bamfdaemon
 16.2 mW    687,7 µs/s       4,0        Process        /usr/bin/gtk-window-decorator
 15.1 mW      3,3 ms/s       3,4        Process        /usr/lib/x86_64-linux-gnu/hud/hud-service
 14.5 mW    391,1 µs/s       3,6        Interrupt      [1] i8042
 10.9 mW    248,1 µs/s       2,7        Interrupt      [4] block(softirq)

Why if I sum all the values in the Power est. column I don't get the 38.1W of the discharge rate, but only something like 8~9W?


The battery reports a discharge rate of 38.1 W

That value is read directly from battery/motherboard over ACPI, and should reflect correct power usage relatively well (not on all devices, though). However, if this report from battery is consistently off, estimated power should still match it.

Values in "Power est." column are estimated by PowerTOP, by observing power usage under different loads and display brightnesses. PowerTOP gets baseline power when your display is off and processor (and GPU) is in sleep state. On top of that, there're power estimates.

There is no mechanism to obtain power usage of single component or process. For example, some peripherals could consume remarkable amount of power, but PowerTOP have no way to measure that. It is better to show best possible estimates for impact of each process, and have (relatively large) chunk of unaccounted usage than to scale estimates to match discharge rate.

To improve estimates, you can try (re)calibrating PowerTOP with

powertop --calibrate

From man 8 powertop:

    runs powertop in calibration mode. When running on battery, 
    powertop can track power consumption as well as system activity. 
    When there are enough measurements, powertop can start to report 
    power  estimates.  One can get more accurate estimates by using 
    this option to enable a calibration cycle. This will cycle through 
    various display levesl (sic) and USB device activities and workloads.

Try running calibration with unplugged charger. There's small chance that either your laptop reports battery charging status incorrectly, or you hit a bug in PowerTOP.

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.