This document presents a cache hierarchy-aware task scheduling (CHATS) algorithm designed to enhance cache performance in parallel loops on chip-multiprocessor (CMP) architectures. CHATS optimizes spatial and temporal data locality while maintaining load balancing, achieving up to 25% execution speedup compared to existing scheduling strategies like OpenMP, TBB, and Cilk++. The study evaluates CHATS against various scheduling policies using synthetic and real workloads, demonstrating its effectiveness in improving cache utilization and execution efficiency.