Title :
Dataflow software pipelining: a case study
Author :
Gao, Guang R. ; Paraskevas, Z.
Author_Institution :
Sch. of Comput. Sci., McGill Univ., Montreal, Que., Canada
Abstract :
Summary form only given. The authors report a case study of data-flow software pipelining, an efficient code-mapping strategy for array operations in loops on a highly pipelined static data-flow processor architecture based on an argument-fetching data-driven principle. The new architecture has the potential of keeping the instruction-processing pipeline full busy as long as the structure of the program can keep enough enabled instructions for concurrent execution-one main objective of data-flow software pipelining proposed in this work
Keywords :
parallel architectures; pipeline processing; argument-fetching data-driven principle; array operations; code-mapping strategy; concurrent execution; data-flow software pipelining; enabled instructions; highly pipelined static data-flow processor architecture; instruction-processing pipeline; loops; Computer aided software engineering; Computer architecture; Computer science; Data flow computing; Fires; Parallel processing; Pipeline processing; Processor scheduling; Signal processing; Testing;
Conference_Titel :
Computers and Communications, 1990. Conference Proceedings., Ninth Annual International Phoenix Conference on
Conference_Location :
Scottsdale, AZ
Print_ISBN :
0-8186-2030-7
DOI :
10.1109/PCCC.1990.101718
