Getting performance from AWS ParallelCluster for government customers

Brendan Bouffler
Date : September 22, 2023
Categories : AI/ML ,AWS Batch ,AWS ParallelCluster ,Amazon NICE DCV ,Elastic Fabric Adapter ,Life Sciences
Tags : ai ml ,aws ,batch ,cpus ,dcv ,ec2 ,efa ,gpus ,hpc ,high performance computing ,lustre ,mpi ,nccl ,parallel cluster ,schedulers ,storage ,autoscaling ,aws batch ,bioinformatics ,cloud computing ,elastic ,elastic fabric adapter ,government ,govt ,hpc instances ,infiniband ,job scheduling ,scientific computing ,supercomputing ,technical computing ,tightly coupled ,virtualization ,vizualization ,techshorts

If you’re pondering a new procurement phase to build an HPC facility for your lab or agency, it’s worth zooming out for a few minutes and seeing if the things you think you know about the cloud are still true. A lot has happened - especially at AWS in the last few years.

This is the fifth show in a multi-part series to bring you up to speed with HPC at AWS - from a government customer point of view.

Today we talk about performance aspects of ParallelCluster - from networking and GPUs to placement groups and everything in between.

If you have ideas for technical topics you’d like to see us cover in a future show, let us know by emailing us at ask-hpc@amazon.com.