Evaluation of static and dynamic task mapping algorithms in NoC-based MPSoCs

Abstract

Task mapping is an important issue in MPSoC design. Most recent mapping algorithms perform them at design time, an approach known as static mapping. Nonetheless, applications running in MPSoCs may execute a varying number of simultaneous tasks. In some cases, applications may be defined only after system design, enforcing a scenario that requires the use of dynamic task mapping. Static mappings have as main advantage the global view of the system, while dynamic mappings normally provide a local view, which considers only the neighborhood of the mapping task. This work aims to evaluate the pros and cons of static and dynamic mapping solutions. Due to the global system view, it is expected that static mapping algorithms achieve superior performance (w.r.t. latency, congestion, energy consumption). As dynamic scenarios are a trend in present MPSoC designs, the cost of dynamic mapping algorithms must be known, and directions to improve the quality of such algorithms should be provided without increasing execution time. This quantitative comparison between static and dynamic mapping algorithms is the main contribution of this work.

Dynamic Resource Allocation in Embedded, High-Performance and Cloud Computing

“…A multi-task dynamic mapping approach was proposed in [128]. Singh et al [128] extends the work described in [32], which evaluates the power dissipation as the product of number of bits to be transferred and distance between source-destination pair.…”

“…P-State conditional increase (lines [30][31][32][33]: If no new tasks have been fetched from the task queue for time Γ and the processor performance is not the lowest possible (P-State is different from P max ), the processor P-State is increased (line 31) and the current time is assigned to γ (line 33).…”

Dynamic task mapping for Network-on-Chip based systems

“…Carvalho et al [6] conducted quantitative comparison to find out the pros and cons of static and dynamic mapping algorithms. According to the results of the study, dynamic algorithms are more advantageous than static techniques as they are able to deal with real-time scenarios.…”

“…Application or task mapping is one of the critical factors that affect the performance and efficiency of NoC based MPSoCs. There are few surveys and comparative studies conducted to evaluate the task mapping algorithms proposed for NoCs [4,5,6]. However, these studies either restricted the analysis to smaller mesh sizes or have conducted qualitative comparison of presented algorithms.…”

“…However, these studies either restricted the analysis to smaller mesh sizes or have conducted qualitative comparison of presented algorithms. For instance, in [6] authors have conducted the comparison of static and dynamic task mapping heuristics for 5×5 and 9×9 NoCs. Similarly, the authors in [5] have also presented the analysis of selected dynamic task mapping heuristics over 8×8 mesh NoCs.…”