14 DAY TRIAL // Amazon has hit the mark with its series of cloud computing services. Not exactly the first company on most people’s minds when thinking about performance computing a few years ago, the company has made a name for itself by offering low-cost, flexible computing power to those who needed it. Now it’s expanding its offering with a more targeted product called Cluster Compute, based on its popular Elastic Cloud Computing (EC2) service. The new offering is aimed at those needing higher performance than what EC2 usually offers.
“Businesses and researchers have long been utilizing Amazon EC2 to run highly parallel workloads ranging from genomics sequence analysis and automotive design to financial modeling. At the same time, these customers have told us that many of their largest, most complex workloads required additional network performance," Peter De Santis, General Manager of Amazon EC2, said.
"Cluster Compute Instances provide network latency and bandwidth that previously could only be obtained with expensive, capital intensive, custom-built compute clusters,” he added.
Typically, companies that needed this kind of computing power had to build their own high-performance computing clusters. These can be prohibitively expensive and only the largest companies can afford them.
With Amazon’s Cluster Compute Instances, customers are guaranteed that all of the computing nodes used for their workload are physically close together and linked through 10 Gbps network connections. This is especially useful for tasks that tasks that run a large number of connected parallel processes.
Those who have tested Amazon’s new offering say there are cases where applications perform as much as 8.5 times better than with the regular EC2 instances. Other than the emphasis on lower latency and higher performance, Cluster Compute instances will be very much like the regular offering and will be managed in exactly the same way. At launch, a Cluster Compute instance will be powered by two Intel Xeon X5570 processors and prices will start at $1.60 per hour.