CO2 cost of Google searches
12/01/09 09:07 Filed in: Internet
Making two internet searches through Google produces
about the same amount of carbon dioxide as boiling a
kettle, claims
a story in the Telegraph.
Mmm...Interesting story but as normal, it seems, with a story about technology they miss a vital point.
Mmm...Interesting story but as normal, it seems, with a story about technology they miss a vital point.
Yes doing web searches does produce CO2, there is no
doubt about that fact.
However not doing web searches also produces CO2!
Okay how do I come to that conclusion? Very simple really, googles servers are running all of the time, they have to do be. A server’s job is to serve so it has to be ready and waiting for those requests all the time. A single search request to google, which will probably be serviced by many servers in fact, will increase the load on these servers ever so slightly. I guess that the increase in load is tiny compared with the base load of the server.
I’m guessing this how google get to the figure mentioned in the article of 0.2g of CO2 pre request, they take the average power consumption of the server and divide it by the average number of requests it handles.
But even this extra complexity doesn’t cover the whole picture.
If you think about it, if there where no search engines you would maybe have to visit 100’s of web sites to find what you want, rather than the 5 or 10 that it normally takes after a search. Each one of these visits would add, slightly to the load of a web server somewhere. Now for a site to show on google or in deed any search engine it needs to have been visited by the engines spiders at least once, but that visit by the spider could save hundreds of visits to the site which don’t result in found content.
So searching does produce CO2, but probably a lot less than not searching.
And as was mentioned in the article google do make large efforts to reduce their data centers power consumption, it makes sense for them to do so, I would guess power is one of the biggest costs (second only to bandwidth I imagine). They have servers specially made that don’t have video cards in them (they are never connected to monitors so why bother?), they use evaporators instead of condensers to cool their data centers. Other search providers may well do similar things as well, but I suspect google got there first simply because their need to save energy was the biggest.
So yes, searching produces CO2, we have to accept that. The search providers can do things about reducing this, but while we still generate electricity by burning stuff this is inescapable, but google are helping here as well, with there RE program.
However not doing web searches also produces CO2!
Okay how do I come to that conclusion? Very simple really, googles servers are running all of the time, they have to do be. A server’s job is to serve so it has to be ready and waiting for those requests all the time. A single search request to google, which will probably be serviced by many servers in fact, will increase the load on these servers ever so slightly. I guess that the increase in load is tiny compared with the base load of the server.
I’m guessing this how google get to the figure mentioned in the article of 0.2g of CO2 pre request, they take the average power consumption of the server and divide it by the average number of requests it handles.
But even this extra complexity doesn’t cover the whole picture.
If you think about it, if there where no search engines you would maybe have to visit 100’s of web sites to find what you want, rather than the 5 or 10 that it normally takes after a search. Each one of these visits would add, slightly to the load of a web server somewhere. Now for a site to show on google or in deed any search engine it needs to have been visited by the engines spiders at least once, but that visit by the spider could save hundreds of visits to the site which don’t result in found content.
So searching does produce CO2, but probably a lot less than not searching.
And as was mentioned in the article google do make large efforts to reduce their data centers power consumption, it makes sense for them to do so, I would guess power is one of the biggest costs (second only to bandwidth I imagine). They have servers specially made that don’t have video cards in them (they are never connected to monitors so why bother?), they use evaporators instead of condensers to cool their data centers. Other search providers may well do similar things as well, but I suspect google got there first simply because their need to save energy was the biggest.
So yes, searching produces CO2, we have to accept that. The search providers can do things about reducing this, but while we still generate electricity by burning stuff this is inescapable, but google are helping here as well, with there RE program.
|