less power when they arent actually doing anything, even with power saving
features such as frequency scaling turned off. Furthermore, processors
dont actually use that much power. The highest I could find on this page
http://en.wikipedia.org/wiki/CPU_power_dissipation
was 130W. Computer power supplies are generally sized so that they dont
explode when you are burning two DVD's with the CPU maxed out. That means
the power supply never actually sees power being drawn at its maximum
rating of 350W. So no, we wouldn't need to add 350GW of power production.
A typical desktop will draw about 60W when idle, and a typical laptop will
draw 25W at idle. Let's face it, most computers are idle most of the time.
Both types of computers draw near zero when in sleep or standby mode. Many
computers are not configured by default to use sleep mode, and people are
not going to figure out how to enable it. Then there is the short delay in
starting up which leads many people to turn off this feature. A computer
could "learn" its user's habits so that it will already be started up by
the time they wanted to use it. So, this is a software problem really.
On Wed, 10 Oct 2007, Simón Ruiz wrote:
> Date: Wed, 10 Oct 2007 16:52:01 -0400
> From: "[ISO-8859-1] Simón Ruiz" <simon.a.ruiz@gmail.com>
> Reply-To: blug@cs.indiana.edu
> To: blug@cs.indiana.edu
> Subject: Re: [BLUG] NOV meeting topic
>
> At the OLF, maddog gave a presentation about computing and power
> consumption. Interesting numbers game: At 350 watts per computer, in
> order to double the number of personal computers on the planet
> (1,000,000,000), we would need to build 25 power plants with output
> equal to the single largest power plant we have today.
>
> As a geek with hippy tendencies, this has made me start thinking about
> how to make computing more earth-friendly. There are quite a few
> projects/business ventures going on right now that are working on
> that. The OLPC for one, and the nifty little Koolu boxes for another.
>
> I think the ideal would probably be ultra-power-efficient massive
> back-end servers and solid-state, fanless, maybe even PoE-fed
> thin-client front ends.
>
> The other thing I see as a big possibility would be to have beefier
> thin-client front-ends that contribute to the processor requirements
> of the whole system through an OpenMosix-style back-end pool of
> processor time.
>
> So in a building of 300 connected workstations, the whole building
> would have 300 workstations worth of processor power. If you're only
> using a word processor or a browser or an e-mail client or something,
> your idle processor time would be up for someone else's use. But say
> you're trying to render 3-d graphics or fold proteins or something
> that can use as much processor time as it can get, you could then
> start enlisting the idle processor time of all your neighbors for your
> task.
>
> What do you think? Silly or workable?
>
> Simón
>
> On 10/10/07, jtwelty@indiana.edu <jtwelty@indiana.edu> wrote:
>> According to the specs, the typical processor power consumption is 95
>> watts; up to 123 watts... pretty efficient.
>>
>> Quoting Simón Ruiz <simon.a.ruiz@gmail.com>:
>>
>>> Out of curiosity, how expensive is this set up, and how much power
>>> does it draw?
>>>
>>> On 10/9/07, jtwelty@indiana.edu <jtwelty@indiana.edu> wrote:
>>>> Sun Blade T6320 with 8 cores - 64 threads, 64 GB memory, 176 Gbps I/O
>>>> throughput, 10 GbE networking.
>>> _______________________________________________
>>> BLUG mailing list
>>> BLUG@linuxfan.com
>>> http://mailman.cs.indiana.edu/mailman/listinfo/blug
>>>
>>
>>
>>
>>
>> _______________________________________________
>> BLUG mailing list
>> BLUG@linuxfan.com
>> http://mailman.cs.indiana.edu/mailman/listinfo/blug
>>
>
> _______________________________________________
> BLUG mailing list
> BLUG@linuxfan.com
> http://mailman.cs.indiana.edu/mailman/listinfo/blug
>
