Amazon Elastic Compute Cloud (EC2) was announced on the AWS blog on August 25, 2006, by Jeff Barr, as a limited beta. The launch post described virtual computing environments specified at the time as a 1.7 GHz Xeon processor, 1.75 GB of RAM, 160 GB of local disk, and 250 Mb/second of network bandwidth, available at 10 cents per hour.
EC2 was built on the Xen hypervisor, and its servers booted from Amazon Machine Images (AMIs), pre-configured boot disks stored as objects in S3. During the beta, an account could run up to 20 virtual servers at a time. The combination of EC2 for compute and S3 for storage gave developers a complete, rentable infrastructure accessible entirely over the web.
The significance of EC2 is that it made “the cloud” concrete for working developers. As the launch post emphasized, developers no longer had to purchase hardware in advance; they could start, stop, and scale virtual servers in response to actual demand and pay only for the hours they ran. This was particularly economical for variable workloads such as seasonal processing or traffic spikes.
Together with S3, launched five months earlier, EC2 established the model of on-demand, metered infrastructure that came to define public cloud computing.