Traditional parallel compilers do not effectively parallelize ir-regular applications because they contain little loop-level parallelism. We explore Speculative Task Parallelism (STP), where tasks are full proce-dures and entire natural loops. Through profiling and compiler analysis, we find tasks that are speculatively memory-and control-independent of their neighboring code. Via speculative futures, these tasks may be executed in parallel with preceding code when there is a high probability of independence. We estimate the amount of STP in irregular appli-cations by measuring the number of memory-independent instructions these tasks expose. We find that 7 to 22% of dynamic instructions are within memory-independent tasks, depending on assumptions.
CITATION STYLE
Kreaseck, B., Tullsen, D., & Calder, B. (2000). Limits of task-based parallelism in irregular applications. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1940, pp. 43–58). Springer Verlag. https://doi.org/10.1007/3-540-39999-2_6
Mendeley helps you to discover research relevant for your work.