THE USE OF TASK PRIORITIZATION IN OPENMP

As the hardware is evolving, the amount of cores increases and so does the use of parallel computing. Parallelizing a program is a way to increase the performance by distributing tasks across multiple threads or processors, taking advantage of the increasing amount of cores. An application programming interface that simplifi\u80es the usage of parallelism is OpenMP. OpenMP uses environment variables to in\u83uence the runtime behavior and one such variable that will be analized in this study is OpenMP 4\u945 Task Priority. Th\u91is study analysis how Task Priority can be used for effi\u81ciently prioritizing computations to increase the execution time, and also the limitations of Task Priority. \u91The study shows that Task Priority can improve the performance of a program by prioritizing the execution of some tasks before others