Macro-actor execution on multilevel data-driven architectures

Conference · Sat Dec 31 00:00:00 EST 1988

OSTI ID:10191989

Gaudiot, J L ^[1]; Najjar, W ^[2]

---

1. University of Southern California, Los Angeles, CA (United States). Computer Research Inst.
2. University of Southern California, Marina Del Rey, CA (United States). Information Sciences Inst.

The data-flow model of computation brings to multiprocessors high programmability at the expense of increased overhead. Applying the model at a higher level leads to better performance but also introduces loss of parallelism. We demonstrate here syntax directed program decomposition methods for the creation of large macro-actors in numerical algorithms. In order to alleviate some of the problems introduced by the lower resolution interpretation, we describe a multi-level of resolution and analyze the requirements for its actual hardware and software integration.

• Research Organization: University of Southern California, Los Angeles, CA (United States). Computer Research Inst.
• Sponsoring Organization: USDOE, Washington, DC (United States); Department of Defense, Washington, DC (United States)
• DOE Contract Number: FG03-87ER25043
• OSTI ID: 10191989
• Report Number(s): CONF-8804338-1; ON: DE94000318; CNN: Contract N00014-86-K-0311
• Resource Relation: Conference: 1. Italian workshop on parallel architectures and neural networks,Salerno (Italy),27-29 Apr 1988; Other Information: PBD: [1988]
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
ARRAY PROCESSORS
DATA-FLOW PROCESSING
COMPUTER ARCHITECTURE
PERFORMANCE
PARALLEL PROCESSING
ALGORITHMS
990200
MATHEMATICS AND COMPUTERS