Can turning Google's background black save power? Hardly.
- Google's response to the theory
The official Google blog confirms their belief in independent studies showing no distinct power savings from a color or layout switch.
- An initial test
This is a smaller test to see if Blackle's claim had any credibility.
- Ways to save energy that work
This is an awesome list of things to do to conserve energy (sorry, no easy shortcuts!)
Blackle is a website that utilizes a customized Google search bar. The website asserts that it is exactly like Google except that it uses a black background instead of a light colored background. This first claim is questionable.
The second claim is that a computer monitor requires more energy to display a white pixel than a black pixel. This amount of energy is minimal, as admitted by the Blackle authors. They insist that there would be significant energy savings if all the millions of screens that look at Google every day used Blackle instead.
Firstly, color is not the only difference between Google and Blackle. Any search through Blackle uses Blackle's servers before querying Google's servers. That means a little extra bit of bytes being transferred. It's not much, but neither is the energy savings.
Secondly, home studies suggest that LCD monitors have a negligible power savings between displaying dark screens or light screens. I say negligible because the variance from one monitor to another is a wider spread of kilowatt-hours than the savings on any particular monitor.
For more details and to see the science (which is the cool and important part) see the links section.
Why lie or exaggerate the benefits of Blackle?
Every time a person uses Blackle, Google shares potential revenue from programs like AdWords with the Blackle site owners. By routing thousands of users through Blackle under the pretense of energy conservation, the Blackle team gets a big cash bonus in their AdSense account.
Anybody with an AdSense account can make money in this way.
The frontpage counter shows how many kilowatt-hours have been saved by using Blackle even though this is an impossible calculation. There are variables which are simply unattainable from Blackle's point of view so this is by necessity only an approximation of the energy saved assuming everybody is using the same monitor that saves the same amount of energy while the Blackle screen is up. Furthermore, it assumes the Blackle screen is maximized even though the counter probably continues at the same pace regardless of how much of my monitor is occupied by the Blackle window.
Incidentally, it may be unfair to say they're lying. It would be more apt to say the site exaggerates the benefits of the Black display color, misrepresents the Roberson et al study, and provides a front page approximation as if it were literal fact.
CRT vs. LCD
Energy savings for CRT monitors displaying white and black screens might still be significant. But an LCD monitor displaying any color is using far less power than a CRT monitor displaying a dark screen. If you're concerned about energy savings, use LCD over CRT. Also, turn off your monitor when you're not using it.
While we're at it, we might as well mention that power management is useful, but screensavers are not. Also, Unplugging things from the wall is even better than turning it off since the plug still drains energy even if the device is not on.
Roberson et al
The Roberson et al study that Blackle cites in its About section specifically states on page 7 that over half of the computer monitors used in the test were CRT monitors. This is a strange choice since LCD monitors dominate three quarters of the market (despite the fact that they have been commerically available for less than half the amount of time that CRT monitors have been).
Page 19 concedes "it appears that display color is a significant determinant of on power for CRTs, but not for LCDs." So this is pretty open and shut exaggeration on Blackle's part. To imply that we can all save energy by using Blackle is dangerously close to lying when the report Blackle cites clearly says that LCD monitors (which most of us have) show no significant change between colors.
The study is well performed, but the sample size is as limited as the studies I link to at the top of this page. And the sampling is not representative of the global average of Blackle or Google users.