MONITORING SYSTEM DESIGN IN HIGH PERFOMANCE COMPUTING CENTER FOR SOLVING MATERIAL SCIENCE PROBLEMS

The paper considers methods for monitoring the workload of resources of a high-performance complex when setting computational problems of materials science in a common execution queue. Since materials science algorithms can be effectively run in parallel and are demanding on computing resources, there is a need to use supercomputer systems. However, inefficient use of resources significantly increases the estimated time of tasks. We propose to implement a monitoring system for subsequent analysis of allocated computing resources.