Loading...
Please wait, while we are loading the content...
Similar Documents
A New Algorithm for Global Optimization for Parallelism and Locality (0)
| Content Provider | CiteSeerX |
|---|---|
| Author | Appelbe, Bill Doddapaneni, Srinivas Hardnett, Charles |
| Description | . Converting sequential programs to execute on parallel computers is difficult because of the need to globally optimize for both parallelism and data locality. The choice of which loop nests to parallelize, and how, drastically affects data locality. Similarly, data distribution directives, such as DISTRIBUTE in High Performance Fortran (HPF), affects available parallelism and locality. What is needed is a systematic approach to converting programs to parallel form, based upon analysis that identifies opportunities for both parallelism and locality in one representation. This paper presents a global framework for optimizing parallelism and locality, based upon constraint solving for locality between potentially parallel loop nests. We outline the theory behind the framework, and provide a global algorithm for parallelizing programs while optimizing for locality. We also give results from applying the algorithm to parallelizing the Perfect benchmarks, targeted at the KSR-1, and analyze ... In Proc. 7th Ann. Workshop on Languages and Compilers for Parallel Computing |
| File Format | |
| Language | English |
| Access Restriction | Open |
| Subject Keyword | Perfect Benchmark Global Algorithm Global Optimization Parallel Loop Nest Data Distribution Directive New Algorithm Parallel Computer Global Framework Systematic Approach High Performance Fortran Sequential Program Loop Nest Available Parallelism Data Locality |
| Content Type | Text |
| Resource Type | Article |
