DocumentCode
128883
Title
Program affinity performance models for performance and utilization
Author
Moore, Ryan W. ; Childers, Bruce R.
Author_Institution
Comput. Sci. Dept., Univ. of Pittsburgh Pittsburgh, Pittsburgh, PA, USA
fYear
2014
fDate
24-28 March 2014
Firstpage
1
Lastpage
4
Abstract
Multithreaded applications have a wide variety of behavior, causing complex interactions with today´s chip multiprocessor machines. Application threads may have large private working sets, and may compete for cache space and memory bandwidth. These threads benefit from large private caches. Other threads may share data or communicate, and thus, execute more quickly if using shared caches. Many applications fall somewhere in between, requiring careful thread-to-core assignments to maximize performance. Yet because of the large number of thread-to-core assignments on today´s chip multiprocessors, it is time and energy prohibitive to exhaustively try and determine the best assignment. In this paper, we present and demonstrate application performance models that predict application performance given a proposed thread-to-core assignment. We show how these models can be quickly built and used to select thread-to-core assignments for multiple programs and to improve system utilization.
Keywords
cache storage; microprocessor chips; multiprocessing systems; application performance; application threads; chip multiprocessor machines; complex interactions; multiple programs; multithreaded applications; private caches; program affinity performance; shared caches; system utilization; thread-to-core assignments; Bandwidth; Computational modeling; Instruction sets; Message systems; Predictive models; Sockets; Training;
fLanguage
English
Publisher
ieee
Conference_Titel
Design, Automation and Test in Europe Conference and Exhibition (DATE), 2014
Conference_Location
Dresden
Type
conf
DOI
10.7873/DATE.2014.036
Filename
6800237
Link To Document