Case Study: National Computational Infrastructure achieves scalable, high-performance workload management for Raijin supercomputer with Altair PBS Professional

PBS Professional Manages Workload for NCI Raijin, Largest Supercomputer in Southern Hemisphere

National Computational Infrastructure (NCI), Australia’s national research computing service, needed a highly scalable, flexible workload manager for its Fujitsu PRIMERGY “Raijin” supercomputer to handle a wide spectrum of job types and replace a difficult-to-maintain in-house OpenPBS system. After a competitive bake-off, NCI selected Altair’s PBS Professional to meet demanding performance, customization and accounting requirements.

Altair implemented PBS Professional and provided Professional Services to integrate with OneSIS and OpenStack, build a plugin-based accounting/granting system, and train NCI staff. PBS Professional now runs in production on Raijin’s 57,472 Intel cores (≈1.2 PFlops) with a 9 PByte Lustre scratch, serving over 1,000 users and managing compute, storage and cloud resources—delivering measurable scalability, flexibility and improved resource accounting for NCI.

National Computational Infrastructure

Ben Evans

Associate Director