POLLY — PERFORMING POLYHEDRAL OPTIMIZATIONS ON A LOW-LEVEL INTERMEDIATE REPRESENTATION

Author:

GROSSER TOBIAS1,GROESSLINGER ARMIN2,LENGAUER CHRISTIAN2

Affiliation:

1. Parkas Group, Computer Science Department, École Normale Supérieure/INRIA, 45 Rue d'Ulm, Paris, 75005, France

2. Programming Group, Department of Informatics and Mathematics, University of Passau, Innstraße 33, Passau, 94032, Germany

Abstract

The polyhedral model for loop parallelization has proved to be an effective tool for advanced optimization and automatic parallelization of programs in higher-level languages. Yet, to integrate such optimizations seamlessly into production compilers, they must be performed on the compiler's internal, low-level, intermediate representation (IR). With Polly, we present an infrastructure for polyhedral optimizations on such an IR. We describe the detection of program parts amenable to a polyhedral optimization (so-called static control parts), their translation to a Z-polyhedral representation, optimizations on this representation and the generation of optimized IR code. Furthermore, we define an interface for connecting external optimizers and present a novel way of using the parallelism they introduce to generate SIMD and OpenMP code. To evaluate Polly, we compile the PolyBench 2.0 benchmarks fully automatically with PLuTo as external optimizer and parallelizer. We can report on significant speedups.

Publisher

World Scientific Pub Co Pte Lt

Subject

Hardware and Architecture,Theoretical Computer Science,Software

Cited by 219 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. IRIS: A Performance-Portable Framework for Cross-Platform Heterogeneous Computing;IEEE Transactions on Parallel and Distributed Systems;2024-10

2. Abstractions for C++ code optimizations in parallel high-performance applications;Parallel Computing;2024-09

3. Fusing Gathers with Integer Linear Programming;Proceedings of the 1st ACM SIGPLAN International Workshop on Functional Programming for Productivity and Performance;2024-08-28

4. A Survey of General-purpose Polyhedral Compilers;ACM Transactions on Architecture and Code Optimization;2024-06-22

5. (De/Re)-Composition of Data-Parallel Computations via Multi-Dimensional Homomorphisms;ACM Transactions on Programming Languages and Systems;2024-05-22

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3