Paper Info
Title
Automatic Recognition of Performance Idioms in Scientific Applications
Abstract
Basic data flow patterns that we call performance idioms, such as stream, transpose, reduction, random access and stencil, are common in scientific numerical applications. We hypothesize that a small number of idioms can cover most programming constructs that dominate the execution time of scientific codes and can be used to approximate the application performance. To check these hypotheses, we proposed an automatic idioms recognition method and implemented the method, based on the open source compiler Open64. With the NAS Parallel Benchmark (NPB) as a case study, the prototype system is about 90% accurate compared with idiom classification by a human expert. Our results showed that the above five idioms suffice to cover 100% of the six NPB codes (MG, CG, FT, BT, SP and LU). We also compared the performance of our idiom benchmarks with their corresponding instances in the NPB codes on two different platforms with different methods. The approximation accuracy is up to 96.6%. The contribution is to show that a small set of idioms can cover more complex codes, that idioms can be recognized automatically, and that suitably defined idioms may approximate application performance.
Year
DOI
Venue
2011
10.1109/IPDPS.2011.21
IPDPS
Keywords
Field
DocType
public domain software,approximation accuracy,scientific application,automatic idioms recognition,application performance,idiom benchmarks,automatic recognition,npb code,parallel benchmark,prototype system,idiom classification,different platform,natural sciences computing,approximate application performance,automatic idioms recognition method,open source compiler open64,performance idioms,data flow pattern,complex codes,different method,abstract data types,performance idiom,benchmark testing,scientific code,scientific codes,scientific applications,program compilers,indexes,data flow,optimization,prototypes,random access,liquefied natural gas,indexation,hardware
Programming language,Computer science,Stencil,Theoretical computer science,Small set,Benchmark (computing),Data flow diagram,Distributed computing,Abstract data type,Transpose,Parallel computing,Compiler,Random access
Conference
ISSN
ISBN
Citations 
1530-2075 E-ISBN : 978-0-7695-4385-7
978-0-7695-4385-7
7
PageRank 
References 
Authors
0.51
15
4
Name
Order
Citations
PageRank
Jiahua He11499.28
allan snavely a2117770.79
Rob F. Van der Wijngaart337445.61
Michael A. Frumkin412619.68