Download:
|
by Jennifer M. Anderson, Mary W. Hall, Saman P. Amarasinghe, Brian R. Murphy, Shih-wei Liao, Edouard Bugnion, Monica S. Lam
http://www.cs.umd.edu/class/spring1999/cmsc732/papers/suif-dec98.ps
Add To MetaCart
Abstract:
The SUIF compiler automatically parallelizes and optimizes sequential programs for shared memory multiprocessors. This paper focuses on new techniques in SUIF for locating coarse-grain parallelism and for managing multiprocessor memory behavior. Locating coarse-grain parallelism gives the individual processors large units of independent computation and reduces the amount of synchronization and communication. Managing the memory behavior involves improving the temporal and spatial locality of the application to reduce cache misses. The SUIF compiler was developed at Stanford University and supports Alpha-based systems. We applied SUIF to two suites of scientific applications, the NAS benchmarks and the SPECfp95 benchmarks. We present performance results on eight-processor Digital AlphaServers, and show that SUIF's analyses have a significant impact on the multiprocessor performance of these applications. 1
Citations
|
241
|
Global optimizations for parallelism and locality on scalable parallel machines
– Anderson, Lain
- 1993
|
|
212
|
Maximizing Multiprocessor Performance with the SUIF
– Hall, Anderson, et al.
- 1996
|
|
159
|
Data and computation transformation for multiprocessors
– Anderson, Amarasinghe, et al.
- 1995
|
|
125
|
The ParaScope Parallel Programming Environment
– Cooper, Hall, et al.
- 1993
|
|
101
|
Symbolic Analysis for Parallelizing Compilers
– Haghighat, Polychronopoulos
- 1996
|
|
97
|
The PARADIGM Compiler for Distributed-Memory Message Passing Multicomputers
– Banerjee, Chandy, et al.
- 1994
|
|
84
|
Detecting coarse-grain parallelism using an interprocedural parallelizing compiler
– Hall, Amarasinghe, et al.
- 1995
|
|
75
|
Semantical Interprocedural Parallelization: An Overview of the
– Irigoin, Jouvelot, et al.
- 1991
|
|
75
|
Automatic Data Layout for High Performance Fortran
– Kremer
- 1995
|
|
57
|
Interprocedural Symbolic Analysis
– HAVLAK
- 1994
|
|
49
|
Compiler-directed page coloring for multiprocessors
– Bugnion, Anderson, et al.
- 1996
|
|
30
|
Multiprocessors From a Software Perspective
– Amarasinghe, Anderson, et al.
- 1996
|
|
19
|
Effective automatic parallelization with Polaris
– Blume, Eigenmann, et al.
- 1995
|
|
18
|
The GEM optimizing compiler system
– Blickstein, Craig, et al.
- 1992
|
|
6
|
Performance Evaluation Corporation. Digital Equipment Corporation Alphaserver 8400 5/440 SPEC CFP95 results. spec newsletter
– Standard
- 1996
|
