The Open Cluster Scheduler (OCS) carries forward over three decades of development in HPC workload management – from its origins as the Distributed Queuing System (DQS), through Sun Grid Engine (SGE), to today’s actively developed open-source project.
Gridware Cluster Scheduler (GCS) builds on this proven foundation with enterprise-grade features and professional support.
This talk explores what’s new and what’s next. We present recent additions to OCS and GCS, including modern scheduling capabilities, improved scalability and security, and enhanced usability for today’s heterogeneous HPC environments.
We also look ahead to the integration of AI as a competent advisor for HPC cluster management – helping administrators optimize resource utilization, diagnose scheduling bottlenecks, and make informed decisions based on workload patterns and cluster behavior.
