System, device and method for monitoring operation efficiency of calculation application software

Abstract

The invention provides a system, device and method for monitoring operation efficiency of calculation application software. The system comprises a super computer platform, a calculation application program operation device and a monitoring device, wherein the super computer platform provides calculation resources for the calculation application program operation device and provides hardware resources for monitoring efficiency of the calculation application program operation device for the monitoring device; the calculation application program operation device allows a corresponding calculation application program to operate on a main node and slave nodes on the super computer platform; the monitoring device starts respective monitors of the main node and the slave nodes on the main node and the slave nodes of the super computer platform; according to physical parameters extracted from all the nodes and effective monitoring data, sent by the main node, of all the nodes, the operation efficiency of the calculation application program operation device is calculated. According to the system, device and method, extension and the like of the calculation application program can be determined and enough data index bases can be provided for maximization of application performance.

Description

technical field [0001] The invention relates to a supercomputer and its high-performance computing application software technology, in particular to a system, device and method for monitoring the operating efficiency of large-scale computing application software. Background technique [0002] With the continuous progress of human society and the development of science and technology, people not only have a wider understanding of the natural world, but also have an increasingly urgent need to explore the unknown world. This has led to a sharp increase in the amount of information and data that humans have mastered; at the same time, these massive information and data need to be analyzed and processed in a timely manner. For example, a large astronomical radio telescope array can generate more than 100GB of cosmic microwave data in one second, and these data need to be analyzed in time. For another example, in the field of particle physics research, the data of one collision ...

AI Technical Summary

Problems solved by technology

However, after more than 50 years of development, the current supercomputer architecture has undergone earth-shaking changes. It often has hundreds of thousands or even millions of CPU cores, and a considerable number of supercomputers use hybrid heterogeneous (CPU+GPU / MIC, etc.), which makes the early physical models and mathematical algorithms incapable of doing the job

This is the main reason why the vast majority of application software running on supercomputers is inefficient and poorly scalable

[0004] To overcome the existing technical problems, on the one hand, it is necessary to vigorously research and develop new physical models and mathematical algorithms that match the architecture of today's supercomputers. This is the ultimate means to break through the bottleneck of existing technologies, but it is a Extremely difficult subject, unable to see results and achieve large-scale application in a short period of time

Images