A Compiler-based Communication Analysis Approach for Multiprocessor Systems

In this paper we describe a compiler framework which can identify communication patterns for MPI-based parallel applications. This has the potential of providing significant performance benefits when connec-tions can be established in the network prior to the ac-tual communication operation. Our compiler uses a flexible and powerful communication pattern represen-tation scheme that can capture the property of com-munication patterns and allows manipulations of these patterns. In this way, communication phases can be detected and logically separated within the application. Additionally, we extend the classification of static and dynamic communication patterns and operations to in-clude persistent communications. Persistent communi-cations appear dynamically, however, they remain un-changed for large segments of the application execution. Our compiler is capable of detecting both static and per-sistent communication patterns within an application. We show that for the NAS Parallel Benchmarks, 100% of the point-to-point communications can be classified as either static or persistent and, with the exception of IS, 100 % of the collective were either static or persis-tent. By comparison to application trace data, the pre-dicted LBMHD, CG and MG communication patterns have been verified. 1