LWN.net Logo

Computing power: containing and managing future power requirements (IT-Director)

Bob McDowall discusses computer power consumption issues on IT-Director.com. "Short-term efforts are focussed on energy saving with computer installations. At a simple housekeeping level, for example switching off computers overnight and at weekends, results in energy and cost savings of 70-80%. Equally, switching off your monitor when at lunch, or during periods of absence, can halve the energy consumption." For a broader look at the advantages of power reduction efforts, see Amory Lovins' paper The Negawatt Revolution.
(Log in to post comments)

Does anyone know of an up-to-date survey of efficiency opportunities?

Posted Apr 28, 2006 1:34 UTC (Fri) by xoddam (subscriber, #2322) [Link]

Does anyone know of a recent big-picture survey of efficiency
opportunities? The Lovins paper is 17 years old.

In 1989, 2% of electrical power used in the US went on 'electronics',
probably mostly mainframes and bedside clock radios. Now that every
office desk has a couple of hundred watts of PC burning away day and
night, that proportion will have increased by an order of magnitude as
Amory's 'free' light and heat efficiency improvements have been made.
Though I doubt most of these measures have been carried through -- how
many buildings have ducted daylight to interior spaces? How many windows
are argon-filled IR-reflectors?

The one thing that *has* happened on a large scale is replacement of
incandescent bulbs with compact fluoros, and those don't seem to have
improved much in 15 years. I haven't done it because I hate fluorescent
light (and I have friends who suffer severe adverse reactions to it) --
so I'm probably not using much less power than a similar household would
have done 20 years ago.

The savings to be made through efficiency are still *enormous*. And
every now and then someone stands up and complains that poor California
hasn't built a new power station in 15 years.

Does anyone know of an up-to-date survey of efficiency opportunities?

Posted Apr 28, 2006 13:33 UTC (Fri) by smitty_one_each (subscriber, #28989) [Link]

One wonders what the business plan would look like to have a gym with excercise equipment somehow coupled to generators.
You could trim the fat US population and avoid energy costs simultaneously.

Does anyone know of an up-to-date survey of efficiency opportunities?

Posted May 1, 2006 18:19 UTC (Mon) by XERC (guest, #14626) [Link]

I believe that it wouldn't work. A very simple and rough estimation:

Think of a runner, who runs, let's say, 40min per day, weights about 75kg. Let's suppose that the runner perorms about 2 steps per second(while running). which each step his/her body jumps up and down about 30cm. If we suppose that this applies to the whole 75kg, then we get
75kg*(2 steps)*(0.3meters)*(1 second)=45W. Considering the fact that this, about 40min or, let's say, an hour, per day or per 2 or 3 days, is not that long, then the energy savings were, if the conversion were IDEAL, really rediculously minor. A single fluorecent light bulb is about 36W. Therefore, the gym could be very happy, if it could feed at least halve of its ENERGY SAVING LIGHT BULBS at least SOME OF THE TIME from the energy of their CLIENTS.

I really wonder, if people would start charging the gym for energy that they produce or demand price cuts? ;-)

Does anyone know of an up-to-date survey of efficiency opportunities?

Posted May 2, 2006 1:09 UTC (Tue) by smitty_one_each (subscriber, #28989) [Link]

Sure, the indidividual case is a wash.
Now, parallel it through clever design so that you can combine the efforts of 100 fatbodies, swapping people in and out at will.

Does anyone know of an up-to-date survey of efficiency opportunities?

Posted Apr 28, 2006 13:50 UTC (Fri) by corbet (editor, #1) [Link]

"Does anyone know of a recent big-picture survey of efficiency opportunities? The Lovins paper is 17 years old."

Amory Lovins has certainly not been idle in those 17 years, however. If you're interested in that stuff, look at the work his Rocky Mountain Institute has been doing. There remains quite a bit of low-hanging fruit out there.

Computing power: containing and managing future power requirements (IT-Director)

Posted Apr 28, 2006 19:03 UTC (Fri) by pkern (subscriber, #32883) [Link]

At a simple housekeeping level, for example switching off computers overnight and at weekends, results in energy and cost savings of 70-80%. Equally, switching off your monitor when at lunch, or during periods of absence, can halve the energy consumption.
I guess most Europeans went through enough conditioning to take those measures as granted.

Computing power: containing and managing future power requirements (IT-Director)

Posted May 1, 2006 18:25 UTC (Mon) by XERC (guest, #14626) [Link]

"I guess most Europeans went through enough conditioning to take those measures as granted."

As an european, I can say the following: at home, usually; at work, depends on very many things.

Computing power: containing and managing future power requirements (IT-Director)

Posted May 9, 2006 7:51 UTC (Tue) by shane (subscriber, #3335) [Link]

Agreed. At my last job (in Amsterdam), electricity was included in the cost of the rent, so there was little or no effort spent in reducing power used. In fact, many machines used more power at night, because fancy 3D screen savers would kick on when people left.

Computing power: containing and managing future power requirements (IT-Director)

Posted May 12, 2006 13:42 UTC (Fri) by leandro (guest, #1460) [Link]

switching off computers overnight and at weekends, results in energy and cost savings of 70-80%. Equally, switching off your monitor when at lunch, or during periods of absence, can halve the energy consumption

This is misleading to the point of being counterproductive.

Actually it is just a system policy issue. Either by default configurations in systems installation (mainly for households), or by systems administration (for businesses), computers should turn off screens, suspend and hibernate by inactivity. On one hand, this avoids human errors (forgetting to switch off), avoids productivity loss (having to wait up for booting, set up your sessions etc), and even saves energy and other resources by lengthening the life of equipment, specially monitors, that suffer by repeated switching on and off.

In fact, this big mistake shows one of the biggest problem with losts of ecologists: failure to think at a macro level.

Copyright © 2006, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds