Adaptive concurrency throttling in the exascale era

Abstract

Future multi-core processors will be heterogeneous, be increasingly less reliable, and operate in dynamically changing operating conditions. Such environments will result in a constantly varying pool of hardware resources which can greatly complicate the task of efficiently exposing a program’s parallelism onto these resources. Coupled with this uncertainty is the diverse set of efficiency metrics that users may desire. In this paper we implement Fugu, a runtime library which dynamically, continuously, rapidly and transparently adapts a program’s parallelism, using the Argobots framework.