An Evaluation of Vectorizing Compilers
Authors:
- María J. GarzaránMGMaría J. Garzarán
Top Cited Papers
- 1 October 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 1089795X,p. 372-382
Abstract
Most of today's processors include vector units that have been designed to speedup single threaded programs. Although vector instructions can deliver high performance, writing vector code in assembly language or using intrinsics in high level languages is a time consuming and error-prone task. The alternative is to automate the process of vectorization by using vectorizing compilers. This paper evaluates how well compilers vectorize a synthetic benchmark consisting of 151 loops, two application from Petascale Application Collaboration Teams (PACT), and eight applications from Media Bench II. We evaluated three compilers: GCC (version 4.7.0), ICC (version 12.0) and XLC (version 11.01). Our results show that despite all the work done in vectorization in the last 40 years 45-71% of the loops in the synthetic benchmark and only a few loops from the real applications are vectorized by the compilers we evaluated.Keywords
This publication has 15 references indexed in Scilit:
- Implicit Data Permutation for SIMD Devices2009
- Generating SIMD Vectorized Permutations2008
- Optimizing data permutations for SIMD devices2006
- gprofACM SIGPLAN Notices, 2004
- Design and characterization of the Berkeley multimedia workloadMultimedia Systems, 2002
- A simple framework to calculate the reaching definition of array references and its use in subscript array analysis1999
- A comparative study of automatic vectorizing compilersParallel Computing, 1991
- Data parallel algorithmsCommunications of the ACM, 1986
- Time and Parallel Processor Bounds for Linear Recurrence SystemsIEEE Transactions on Computers, 1975
- An empirical study of FORTRAN programsSoftware: Practice and Experience, 1971