Amazon Web Services Introduces Cluster Compute Instances
Amazon Web Services (AWS) has announced the addition of Cluster Compute Instances for high-performance computing to the Amazon EC2 cloud platform. What this means is that AWS can now handle yet another task normally relegated to on-premises systems. Here are some details.
Cluster Compute Instances are basically what they say on the tin, according to the press release: they provide more CPU power than your average Amazon EC2 instance, and you can group several of them together into clusters to get low latency and high network throughput. Naturally, like all AWS offerings, it’s pay-as-you-go and you can scale as needed.
This is the kind of application that’s been traditionally the province of on-premises computing clusters of the kind you’d find on university campuses or science research facilities. In fact, the press release boasts that The National Energy Research Scientific Computing Center (NERSC) at the Lawrence Berkeley National Laboratory test-drove their high-performance applications with Amazon EC2 Cluster Compute Instances and saw a speed boost of up to 8.5 times.
Now, obviously AWS wouldn’t report numbers that reflected their new product unfavorably. But the fact remains that Cluster Compute Instances are a vast differentiator from other cloud platforms, and opens up a doorway to providing cloud services to a whole new kind of business.