High-performance computing (HPC) for regulated data
High performance computing (HPC) for regulated data
Overview
This service provides a scalable high-performance computing (HPC) environment built on AWS ParallelCluster, with an easy-to-use web interface through Open OnDemand. Jobs are managed with the SLURM scheduler.
Each Principal Investigator (PI) or project has its own dedicated partition that can be customized to meet specific needs, including the type of compute instances, number of cores, amount of memory, and use of GPUs.
Project data is stored on a secure, high performance file system that works like a standard POSIX system. Storage access is managed at the project level and backed by Amazon S3 for durability and reliability.
For Level 3 confidential data or below, refer to high-performance computing for non-regulated data.
Eligibility
Faculty and Researchers in Central Administration, FAS, College, DCE, GSAS, SEAS, GSD, GSE, HBS, HKS, HDS, HLS, HMS (Quad), HSDM (Quad), Radcliffe, SPH
Data classification level
Approved for Level 4 and 5 confidential data