Workflow simulation and multi-threading aware task scheduling for heterogeneous computing

Efficient application scheduling is critical for achieving high performance in heterogeneous computing systems. This problem has proved to be NP-complete even the homogeneous case, heading research efforts obtaining low complexity heuristics that produce good quality schedules. Such an example HEFT, one of most efficient list terms makespan and robustness. In this paper, we propose two task methods systems can integrated several algorithms. First, a method improves time (the output schedule) family algorithms delivered without sacrificing schedule length, when computation costs tasks are unknown. Second, length (makespan) proposed, by identifying which going executed as single-threaded multi-threaded implementations, well number threads used. We showcase both using HEFT popular algorithm, but they other too, such HCPT, HPS, PETS CPOP. The experimental results, consider 14580 random synthetic graphs five real world applications, show enhancing algorithm with proposed methods, significant gains gains, achieved.