Amazon Web Services has announced the availability of Cluster Compute Instances for Amazon EC2, which is a new instance type specifically designed for high-performance computing (HPC) applications and other demanding applications.
The new offering is designing for EC2 clients that have complex computation workloads with large amounts of data, such as financial services. The benefit of Cluster Compute Instances is that it has been engineered to provide more CPU than any other Amazon EC2 instance. And clients can also group Cluster Compute Instances into clusters which allows applications on servers to get the low-latency network performance required for tightly coupled, node-to-node communication.
Amazon says that applications can see up to 10 times the network throughput of the largest current Amazon EC2 instance types. Prior to Cluster Compute Instances for Amazon EC2, customers with advanced HPC would have to implement in-house compute clusters via hardware. The new offering blends AWS’s pay-as-you-go pricing with the ability to scale on demand.
This launch represents Amazon’s continued push to make EC2 stronger and more scalable. And Amazon has continually added new features to all of its AWS offerings. But besides offering customers more features, Amazon may also be prepping for a new competitor. Rumor has it that Google may be launching an EC2 competitor.