Amazon has introduced AWS Parallel Computing Service (PCS), a platform for managing high performance computing (HPC) clusters on AWS EC2. PCS integrates the AWS ParallelCluster software, an open-source HPC cluster management tool introduced by Amazon in 2018, with other tools to fully manage the system for customers who prefer not to maintain their own clusters.
In addition to ParallelCluster, PCS utilizes Slurm as a scheduler to distribute tasks within the cluster and integrates with other AWS services such as the Management Console, Command-Line Interface (AWS CLI), and Elastic File System (EFS). The cost of PCS is based on the number and size of clusters, separate from rental fees for machines and other services that may be applicable. This service is currently available in the Asia Pacific (Singapore) region.
TLDR: Amazon launches AWS Parallel Computing Service for high performance computing clusters on AWS EC2, offering fully managed cluster management for customers, with pricing based on cluster size and usage.
Leave a Comment