1 |
As an example, I am interested in characterizing the power consumption |
2 |
of rendering a PDF document. I would hopefully only need to run the |
3 |
renderer once. |
4 |
|
5 |
I can use PowerTOP, but it seems to be limited to rough measurements |
6 |
on the order of tenths of a watt. This measurement can be divided |
7 |
among the wakeup events in an attempt to calculate software power |
8 |
consumption but it seems imperfect if I want to monitor a single |
9 |
process that may be competing relatively equally for resources with |
10 |
the kernel and other user processes. |
11 |
|
12 |
PMBus is a spinoff of SMBus which is a spinoff of I2C which is found |
13 |
on many motherboards. PMBus is supposed to be the interface which |
14 |
controls and reports power supply activity. Besides the main kW power |
15 |
supply, there is usually a power supply near your processor that steps |
16 |
down 3.3V or 5V to 2.8V, 1.8V, or lower (I've seen as low as 0.8V, but |
17 |
not on a desktop). I was not aware these had a visible interface. |
18 |
|
19 |
Apparently you can talk to these, but my searches can only find code |
20 |
which seems highly experimental. The other replies seem to be for |
21 |
embedded Linux systems running on FPGAs and perhaps Cortex-A parts. |
22 |
|
23 |
If I were using a microcontroller I could get uA or nA draw per MHz |
24 |
and I know my operating voltage and operating time. However, desktop |
25 |
processors are much more complex, and I am not sure if they have been |
26 |
entirely characterized. The most advanced tool I can find is PowerTOP |
27 |
and it does not seem very accurate. |
28 |
|
29 |
Does anyone have any suggestions? Should I start reading source code |
30 |
or post on the forums? Or perhaps someone has used PowerTOP and found |
31 |
it to be reasonably accurate? |
32 |
|
33 |
R0b0t1. |