Amazon doesn’t get a whole lot of love from the “Web 2.0” crowd most of the time — perhaps because it’s kind of a Web 1.0 company that just sells books and other stuff, and happens to use the Web to do it. The giant retailer has added some wiki-type features and other interactivity, but other than that it’s pretty much just a retailer. Bo-ring. Except, of course, for the odd announcement like this one, about Amazon providing what amounts to “grid computing” services — a distributed server network that Web companies can effectively use as their own back-end network.
Nik Cubrilovic, who is a smart guy and runs a Web-based backup company called OmniDrive, has the details at TechCrunch. Apparently (warning: I am not a hardware guy), companies can effectively create a virtual server structured in any way they wish and then upload that image to Amazon’s equipment — that is, its S3 network — and then their service treats that virtual server as though it was just down the hall in a machine room. Users are charged for CPU usage and bandwidth at what appear to be fairly competitive rates (although Nik has some concerns there). We Break Stuff likes it.
This sounds a lot like what Sun Microsystems has been trying to do with its Grid Computing solution, which the struggling server maker — which put the dot in Web 1.0, to paraphrase its famous slogan — launched in 2004, but has apparently had some trouble getting rolling (and getting customers for). By way of a quick comparison, Amazon charges 10 cents per CPU hour for its service, while Sun charges $1 per CPU hour — although I’m sure there are differences of which I am not aware. Sun CEO Jonathan Schwartz wrote a blog post with some background about Sun’s grid efforts back in March.
Update:
There’s a good look at and discussion of Amazon’s EC2 here, and some good comments about its benefits and/or weaknesses at the Reddit page about the announcement here.