Evaluating the Scalability of Dependency-Aware Scheduling in Multicore Systems through Large-Scale Simulation Trials

Task scheduling in multicore systems becomes increasingly complex in the presence of inter-task dependencies, which traditional operating system schedulers are not designed to manage. As a result, tasks are often activated before their prerequisites are satisfied, leading to passive waiting and inefficient CPU utilization. Although several dependency-aware approaches have been proposed, especially in heterogeneous or distributed computing, they often rely on centralized orchestration, static task graph analysis, or dedicated hardware support, limiting their applicability in general-purpose or dynamic environments. Addressing this gap, the Dependency-Aware Model (DAM) has been introduced as a lightweight, user-space scheduling mechanism that delays task activation until all declared dependencies are resolved. To evaluate the scalability and robustness of this model, an extensive empirical campaign was conducted using a dedicated simulation environment capable of executing compute-bound workloads under randomized configurations. The simulator supports variations in thread count, execution time distributions, CPU affinity, and dependency density, while ensuring that task relationships form valid Directed Acyclic Graphs (DAGs). Over 39,000 simulation trials were performed to compare standard and dependency-aware scheduling under diverse conditions. The results consistently demonstrate that DAM improves scheduling efficiency, reducing execution time by over 28% in single-test runs and exceeding 30% in ten-test batches. These findings confirm the model’s practical effectiveness and its suitability for real-world multicore systems with nontrivial dependency structures.