New Screencast: Building a High Performance Cluster
From aeronautics to genomics to financial services, High Performance Computing is becoming a common requirement in many fields of industry and academia. Traditionally, the barrier to entry into this area has remained high, with the expertise and cost needed to provide such facilities proving to be prohibitive.
With Amazon EC2’s Cluster Compute instances, extremely high performance elastic computing is now available in just a few mouse clicks.
With fast network interconnects, high memory and quick CPUs, these instances are extremely capable for tightly coupled tasks or batch processing, and very easy to use. I’ve recorded a short screencast that demonstrates how to build an 8 node, 64 core cluster and kick off a highly parallel analysis run, all in around 10 minutes.