Research Article
Locality-Aware Task Scheduling and Data Distribution for OpenMP Programs on NUMA Systems and Manycore Processors
Figure 8
Thread timelines showing task execution on the eight-node Opteron system. Threads are shown on the -axis and time is shown on the -axis. Memory access latencies are encoded using a green-blue gradient. Tasks stall for fewer cycles under locality-aware scheduling combined with heuristic-guided data distribution.