Amazon Web Services didn’t announce any new price cuts at its annual re:Invent conference last month, but there’s more to the economics of cloud computing than just lower prices. The company demonstrated as much on Tuesday by announcing two new billing tiers for its pre-paid virtual servers, which it calls Reserved Instances.
Now, AWS customers have the option to pay only part of the Reserved Instance cost upfront, or, if they prefer, nothing at all. This is a pretty substantial change from the original all-upfront pricing model in place since AWS launched Reserved Instances in 2009.
The catch with the new payment options is that cost-savings are tied to how much a user pays upfront. According to a blog post announcing the new options, paying the full price of the instances will save users an average of 63 percent off the monthly on-demand price over three years, while paying nothing upfront (an option only available in one-year terms) will result in savings of about 30 percent.
This is just the latest tweak AWS has made to Reserved Instances over the years, previously adding options such as moving them across Availability Zones within AWS regions or letting users change the size of their instances.
Amazon’s announcement also comes just a couple months after [company]Google[/company], which increasingly looks like AWS’s biggest competitor in cloud computing over the next few years, cut its standard compute prices by 10 percent. Last March, Google announced it would calculate billing by the minute rather than by the hour, and that it would offer users reductions of up to 40 percent based on how long a virtual server runs in any given month.