AWS Adds 'Effortless' High Performance Computing
Alces Flight is a new entry in the AWS Marketplace that provides an "effortless" but fully featured High Performance Computing (HPC) environment that can be launched in minutes, according to spokesman Jeff Barr.
"It can make use of On-Demand or Spot Instances and comes complete with a job scheduler and hundreds of HPC applications that are all set up and ready to run," Barr said in a blog post last week. "Some of the applications include built-in collaborative features such as shared graphical views."
EC2 Spot Instances provide spare AWS capacity at up to a 90 percent discount from AWS' On-Demand pricing and can significantly reduce an enterprise's cost per core. Barr said the new HPC offering helps academic and corporate researchers who have been increasingly turning to the cloud to run demanding compute-intensive, highly parallel jobs at a reasonable cost.
They can fire up clusters on demand, run the jobs and shut the clusters down immediately, scaling up clusters or launching new ones as demand grows.
"This self-serve, cloud-based approach favors science over servers and accelerates the pace of research and innovation," Barr said. "Access to shared, cloud-based resources can be granted to colleagues located on the same campus or halfway around the world, without having to worry about potential issues at organizational or network boundaries."
AWS HPC clusters, which behave just like traditional Linux-powered HPC clusters, can launch with fixed sizes or with auto scaling, running in a Virtual Private Cloud (VPC) with SSH and graphical desktop connectivity, while providing access to HPC applications, libraries, tools and other resources.
"You can launch a small cluster (up to 8 nodes) for evaluation and testing or a larger cluster for research," Barr said. "If you subscribe to the product, you can download the AWS CloudFormation template from the Alces site. This template powers all of the products, and is used to quickly launch all of the AWS resources needed to create the cluster."
In his post last week, Barr goes on to detail how to launch an HPC cluster, and the company provides more guidance on a getting started guide.
David Ramel is an editor and writer for Converge360.