Chronus
Wei Gao, Zhisheng Ye, Peng Sun, Yonggang Wen, Tianwei Zhang
Abstract
Modern GPU clusters support Deep Learning training (DLT) jobs in a distributed manner. Job scheduling is the key to improve the training performance, resource utilization and fairness across users. Different training jobs may require various objectives and demands in terms of completion time. How to efficiently satisfy all these requirements is not extensively studied.
Topics & Concepts
Computer scienceKey (lock)Scheduling (production processes)Processor schedulingJob schedulerDistributed computingArtificial intelligenceResource (disambiguation)Cloud computingOperating systemComputer networkOperations managementEngineeringCloud Computing and Resource ManagementAdvanced Neural Network ApplicationsParallel Computing and Optimization Techniques