Cloud computing’s all the rage these days. We’ve all heard the normal spiel about its benefits; cloud services let you reduce your reliance on on-site admins, cloud services let you access data from anywhere, blah blah blah. But did you know that tapping into the cloud for your email services can be up to 80 times more efficient than hosting servers in-house? We didn’t either, until we got our grubby little paws on a new Google report that claimed just that.
Google’s Green Computing: Efficiency At Scale (PDF)
” – mmm, buzzwords – the Goog compared the energy costs of Gmail to the energy costs of in-house enterprise solutions. The study examined data from 4 million businesses that have made the jump to Gmail and found that between server, network and computer energy costs, the average business with an in-house email server consumes waaaaay more energy per user on a yearly basis than a Gmail-running organization. By Google’s count, Gmail users only use 2.2 kWh of energy per year; large businesses with in-house servers use 7.6 kWh per person per year; mid-sized business use 28.4 kWh; while small businesses with on-site email average a whopping 175 kWh of energy per user each and every year.
Google chalks the difference up to the fact that the Gmail servers are almost always fully utilized, they use custom software that was especially designed to produce maximum efficiency, and Google’s only employs high-efficiency servers with high-efficiency power supplies – conditions that most businesses can’t hope to match. On the Google Blog, the Goog’s Program Manager for Green Engineering and Operations, David Jacobowitz,
used a fun analogy to put a Gmail user’s energy use into perspective
“If you’re more of a romantic than a businessperson, think of it this way: It takes more energy to send a message in a bottle than it does to use Gmail for a year, as long as you count the energy used to make the bottle and the wine you drank.”
Of course, Google didn’t take the energy costs of building its high-efficiency servers into account when coming up with their 2.2 kWh number, but it’s still a cool stat.