作者: M.F.P. O'Boyle , Björn Franke
关键词:
摘要: This paper develops a new approach to compiling C programs for multiple address space, multi-processor DSPs. It integrates novel data transformation technique that exposes the processor location of partitioned into parallelization strategy. When this is combined with resolution mechanism, it generates efficient run on spaces without using message passing. applied UTDSP benchmark suite and evaluated four TigerSHARC board, where shown outperform existing approaches gives an average speedup 3.25 parallel benchmarks.