Software Pipeline–Based Partitioning Method with Trade-Off between Workload Balance and Communication Optimization

Abstract

For a multiprocessor System‐on‐Chip (MPSoC) to achieve high performance via parallelism, we must consider how to partition a given application into different components and map the components onto multiple processors. In this paper, we propose a software pipeline–based partitioning method with cyclic dependent task management and communication optimization. During task partitioning, simultaneously considering computation load balance and communication optimization can cause interference, which leads to performance loss. To address this issue, we formulate their constraints and apply an integer linear programming approach to find an optimal partitioning result — one that requires a trade‐off between these two factors. Experimental results on a reconfigurable MPSoC platform demonstrate the effectiveness of the proposed method, with 20% to 40% performance improvements compared to a traditional software pipeline–based partitioning method.

Mapping series-parallel streaming applications on hierarchical platforms with reliability and energy constraints

“…Yu et al [19] proposed a genetic algorithm to map a workflow onto utility grids to minimize execution time under a given budget constraint. In [20], Huang et al partitioned a task graph with cyclic dependencies into parts, and mapped each part to a processor to achieve a balance between communication and workload.…”

Reliable and Energy-aware Mapping of Streaming Series-parallel Applications onto Hierarchical Platforms

“…Yu et al [15] proposed a genetic algorithm to map a workflow onto utility grids to minimize execution time under a given budget constraint. In [16], Huang et al divided the dependent DAG into many parts and mapped each part to the processor to achieve a balance between communication and workload. Wieczorek et al [17] evaluated the performance of the three scheduling strategy for mapping scientific work onto the grid and the experimental results demonstrate that the HEFT algorithm is optimal for balanced and unbalanced applications.…”