Servers as bad as SUV's!

[Gareth Beck]

Hello,

There was an article in data strategy magazine (which I have the pleasure of getting :D ) recently on how the amount of data you hold should be reduced as it has an impact on the enviroment.

As the same theme turned up in New Scientist this week I thought I'd post a link.

New Scientist Data article (http://environment.newscientist.com/channel/earth/dn12992-computer-servers-as-bad-for-climate-as-suvs.html?feedId=online-news_rss20)

Just wondering what peoples thoughts were on this.

Cheers,
Gareth

[Dawg]

The article is just a pot stirrer, no facts no figures, just blather from an organisation that relies for it's funding and existence on this variant of hot air.

I just don't believe that the carbon footprint of a server is the same as that of a SUV. I'd be surprised if it's manufacture footprint was a big as that of a single tyre on a SUV, but if anyone has figures I'm happy to be corrected.

And what of the alternatives? Do we return to storing data on paper? Goodbye trees. Do we not have data, or so little data that it becomes a resource for the rich only? Hello aristocracy of old Etonian environmentalists. (Porritt, Goldsmith etc).

I wish these blatherbimbos wouldn't spout off like this. It just reinforces the deniers, and lessens the impact of the real, credible information about a real, crucial problem.

[ken_uk]

The survey also revealed that considerable amounts of electricity could be saved by more efficient data storage: 60% of the departments said they were using less than half their storage capacity and 37% said they are storing data indefinitely.

If they had said you could save electricity by coding more efficiently then it would be more accurate, as the CPU's are often the most power hungry components in a computer these days. Pushing a cpu to its limits with poorly optimised code will usually consume a lot more power than tightly optimised, efficient code that does not stress the cpu much.

As I see it a half full hard disk drive is not going to use more electricity that a nearly full disk drive, probably the opposite, as one that is ran nearly full all the time is much more likely to be more defragmented than a nice empty one, so consuming more power, as it has to do more disk activity to get the data.

[awebapart.com]

It seems strange to focus on servers when there are a lot more desktops out there, many with similar or higher specs than some servers. Why focus on something that is more difficult to downsize, when downsizing should be easier at the workstation end, after all do you really need a high spec pc to browse the web or send an email?

[Spiderden]

And what of the alternatives? Do we return to storing data on paper? Goodbye trees.

Don't be daft, storing data on a slab of stone with a hammer and chisel is the only green way to do it. ;)

[Pelenna]

From the server rooms I've designed (just the building services, not the technical stuff!), its not the servers themselves that are high consumers, its the huge amount of cooling they require. Not sure exactly how to get around that though without, as you say, reverting back to paper!

[Subbynet]

Does anyone else think the dirty end of power production is now being blamed on all and any device they think is worth targeting?

Put it this way, find a black bag, put your monitor in it, seal it with just the power lead coming out... Then turn it on... Would the bag inflate? Of course it doesn't, because it doesn't produce CO2. Your computer doesn't produce CO2, neither does your TV... No electrical device sits there producing CO2.

What they're doing is shifting the amount of CO2 released from Power Stations and calculating that as directly coming from computers - hence, your problem now.

Overall, all the computers in the world account for 2% of emissions (according to that linked article), but Power Stations which are powering computers account for something like 25% of global emission output from humans.

Basically, 2% is within the tolerance levels set, you can't lower energy production of a Power Station by 2% because its already wasting more than that on a daily basis.

So if we made just a few Power Stations more efficient, or fueled by alternatives, job done, the computers can stay as they're. You can't make "considerable" power savings from computers like that article suggests if it already accounts for a tiny percentage of the power we use.

Carbon Footprint is the biggest con going.

I'm not saying its not worth trying to lower your electricity bill, but thinking it will change the world is wrong.