Partner Certification & Solutions Catalog


Gridware Cluster Scheduler 9.0.x

Gridware Cluster Scheduler (GCS) is an enterprise-grade workload management and job scheduling system designed for high-performance computing (HPC) and AI workloads, built on the open-source Open Cluster Scheduler (OCS) project. GCS delivers exceptional scalability across large compute clusters with high job throughput, while providing flexible resource management and intelligent job prioritization. The system maintains full backward compatibility with Sun Grid Engine and includes native support for MPI, GPU acceleration, and diverse compute architectures including  x86 and ARM.
The 9.0.x series introduces significant performance improvements through a multi-threaded architecture with separate thread pools for different request types, enhanced data stores, and optimized scheduling algorithms for large-scale deployments. This release features improved MPI support for parallel computing workloads, expanded GPU capabilities including support for NVIDIA Grace Hopper and Grace Blackwell architectures, and new flexible resource types for advanced workload management. A preview integration with Prometheus and Grafana enables comprehensive real-time monitoring and visualization of cluster metrics and job performance. The new extensible accounting feature allows administrators to define and track custom job metrics, providing tailored insights into resource usage and workload patterns for maximum performance and observability in HPC and AI environments.
  • Highlights
  • Platform SLES 15
  • Hardware Architecture x86-64
  • Certification SUSE Ready