Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Chaotic linear system solvers in a variable-grain data-driven multiprocessor system

Conference ·
OSTI ID:10192331

Linear systems are important problems in many scientific applications. While asynchronous methods are effective solutions to linear systems, they are difficult to realize due to the chaotic behavior of the algorithms. In this paper, we investigate the implementation as well as the performance of an asynchronous method, namely chaotic relaxation, in our Variable-grain TaggedToken Data-flow (VTD) System. We compare asynchronous methods with synchronous methods executed on both the fine-grain and the coarse-grain execution models. New high-level dataflow language constructs axe introduced in order to express asynchronous operations. A new firing rule that deviates from the single assignment rule of functional languages is proposed to support the implementation of asynchronous computations in the VTD system. In addition to the conventional speedup measure, we then define new performance measurements, called Growth Factor, Scalability Factor, and Robustness to characterize the system performance from the machine and application viewpoints. Simulation results indicate that asynchronous methods axe more efficient than synchronous methods and that the coarse-grain execution mode is more efficient that the fine-grain execution mode in our VTD system.

Research Organization:
University of Southern California, Los Angeles, CA (United States). Dept. of Electrical Engineering
Sponsoring Organization:
USDOE, Washington, DC (United States)
DOE Contract Number:
FG03-87ER25043
OSTI ID:
10192331
Report Number(s):
CONF-9106442--1; ON: DE94000319; BR: KC0701030
Country of Publication:
United States
Language:
English

Similar Records

Solving Partial Differential Equations in a data-driven multiprocessor environment
Conference · Fri Dec 30 23:00:00 EST 1988 · OSTI ID:10190202

Implementing functional languages to exploit locality
Conference · Mon Dec 31 23:00:00 EST 1990 · OSTI ID:5241470

Assessing the benefits of fine-grain parallelism in dataflow programs
Journal Article · Thu Dec 31 23:00:00 EST 1987 · Int. J. Supercomput. Appl.; (United States) · OSTI ID:6524930