Re: Optimizing gcc for power through instruction rescheduling

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Ankit,

You will need extraordinary accuracy and precision in power
measurements to achieve that. Assuming that a quad core processor
dissipates 100 W at 3G, that's 8nW per op. Since it's so small, it
will likely be out-swamped by OS jitter, not to mention the kind
(stability of voltage, frequency) of power supply we usually get in
India.

So, IMHO measuring anything "at the wall" is likely to be unreliable.
However, low power operation in software is a really nice idea. You
could try lesswatts.org. These guys are trying to minimize the power
consumption by reducing the polling frequency of their device drivers
etc. That seems to be a more fruitful venture, at least in the short
term. But hey, many great ideas were laughed off as crazy initially.
I'll be glad to eat my words if you can succeed.

All the best. And do keep me posted.

-- 
Rohit Garg

http://rpg-314.blogspot.com/

Senior Undergraduate
Department of Physics
Indian Institute of Technology
Bombay

[Index of Archives]     [Linux C Programming]     [Linux Kernel]     [eCos]     [Fedora Development]     [Fedora Announce]     [Autoconf]     [The DWARVES Debugging Tools]     [Yosemite Campsites]     [Yosemite News]     [Linux GCC]

  Powered by Linux