English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Today high-performance computing (HPC) platforms are still dominated by batchjobs. Accordingly, effective batch job scheduling is crucial to obtain highsystem efficiency. Existing HPC batch job schedulers typically leverageheuristic priority functions to prioritize and schedule jobs. But, onceconfigured and deployed by the experts, such priority functions can hardlyadapt to the changes of job loads, optimization goals, or system settings,potentially leading to degraded system efficiency when changes occur. Toaddress this fundamental issue, we present RLScheduler, an automated HPC batchjob scheduler built on reinforcement learning. RLScheduler relies on minimalmanual interventions or expert knowledge, but can learn high-quality schedulingpolicies via its own continuous 'trial and error'. We introduce a newkernel-based neural network structure and trajectory filtering mechanism inRLScheduler to improve and stabilize the learning process. Through extensiveevaluations, we confirm that RLScheduler can learn high-quality schedulingpolicies towards various workloads and various optimization goals withrelatively low computation cost. Moreover, we show that the learned modelsperform stably even when applied to unseen workloads, making them practical forproduction use.

RLScheduler: An Automated HPC Batch Job Scheduler Using Reinforcement Learning