Pay-as-you-go AWS Cluster Compute Instances available on-demand
(7/13/2010)
Amazon Web Services, one of the world’s largest providers of public cloud services, has released a new instance type for its Infrastructure-as-a-Service offering that performs high-performance computing.
Cluster Compute Instances for Amazon Elastic Compute Cloud (EC2) is designed for customers with complex workloads, like tightly coupled parallel processes, or applications that are sensitive to network performance. Amazon’s value proposition for cluster compute instances is pay-as-you-go HPC performance on-demand, instead of capital-intensive development of custom HPC infrastructure.
“Cluster Compute Instances provide network latency and bandwidth that previously could only be obtained with expensive, capital-intensive, custom-built compute clusters,” Amazon EC2 General Manager Peter De Santis said in a statement. “For perspective, in one of our pre-production tests, an 880 server sub-cluster achieved 41.82 teraflops on a LINPACK test run.
“We're very excited that Amazon EC2 customers now have access to this type of HPC performance with the low per-hour pricing, elasticity, and functionality they have come to expect from Amazon EC2.”
Researchers at the US Department of Energy’s Lawrence Berkeley National Laboratory test-drove the new offering with positive results. A series of benchmark tests for HPC applications performed by scientists at the lab’s National Energy Research Scientific Computing Center showed that Amazon’s HPC cloud performed 8.5 times faster than the typical EC2 infrastructure providing the company’s regular IaaS offering.
DatacenterDynamics delivers knowledge and networking opportunities to professionals that design, build and operate data centers through a combination of conferences, magazines, websites and research products. Keywords: computer rooms, server farms, datacomm facilities, telecom sites, data centres, colocation, carrier hotels, data center strategy,
