Date: Tue, 14 Jan 2003 17:24:52 -0800 From: "Sam Leffler" <sam@errno.com> To: "Andrew Gallatin" <gallatin@cs.duke.edu>, "Peter Wemm" <peter@wemm.org> Cc: "John Baldwin" <jhb@FreeBSD.org>, "Wilko Bulte" <wkb@freebie.xs4all.nl>, <cvs-committers@FreeBSD.org>, <cvs-all@FreeBSD.org>, "Nate Lawson" <njl@FreeBSD.org> Subject: Re: cvs commit: src/sys/dev/acpica acpi_cpu.c Message-ID: <159a01c2bc34$e8e099b0$5a557f42@errno.com> References: <20030114173224.A74488@grasshopper.cs.duke.edu><20030114230057.81B422A89E@canning.wemm.org> <15908.46788.141735.858350@grasshopper.cs.duke.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
> Peter Wemm writes: > > Andrew Gallatin wrote: > > > John Baldwin [jhb@FreeBSD.org] wrote: > > > > aggressive policies in the future. You could change the speed on > > > > your desktop by changing the performance speed to be 6 (75%) using > > > > the sysctl for example: > > > > > > The scary thing is that for my destop setup, I can turn my speed down > > > to 25% on my P4 and not really feel a difference in xterms, editors > > > and soforth. (I CAN see a difference, I just don't mind it much). > > > > > > Any idea about how much power this actually saves on a desktop? > > > > FWIW, we see a (measured) difference of about 50W on 2.2GHz P4's simply by > > turning machdep.cpu_idle_hlt on and off. I expect the clock throttling > > would make similar differences. For 1U rack-mount systems (especially in > > California) this is a Big Deal. > > Yes, it is. 2 years ago, just as I was leaving Duke and the Enrons of > the world were raping California consumers, some people in my research > group were looking at doing energy-aware server load balancing. Eg, > you suspend a server to save power if load is low, and at peak times > you bring additional servers on line. We were using PIIIs and APM, > waking with wake-on-lan. I bet everything we did would translate to > P4's and ACPI... > > See http://www.cs.duke.edu/ari/publications/balance-of-power.ps > (That's more of a pie-in-the-sky thing.. I don't know where the paper > with the hard numbers went). I believe the paper you're thinking of was given at the 2001 SOSP. There were actually several papers at the conference on the topic. Sam To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?159a01c2bc34$e8e099b0$5a557f42>