Task concurrency management methodology to schedule the MPEG4 IM1 player on a highly parallel processor platform

Abstract

This paper addresses the concurrent task management of complex multi-media systems, like the MPEG4 IM1 player, with emphasis on how to derive energy-cost vs time-budget curves through task scheduling on a multi-processor platform. Starting from the original "standard" specification, we extract the concurrency originally hidden by implementation decisions in a "grey-box" model. Then we have applied two high-level transformations on this model to improve the task-level concurrency. Finally, by scheduling the transformed task-graph, we have derived energy-cost vs time-budget curves. These curves will be used to get globally optimized design decisions when combining subsystems into one complete system or to be used by a dynamic scheduler. The results on the MPEG4 IM1 player confirm the validity of our assumptions and the usefulness of our approach.

A Hardware Implementation of a Run-Time Scheduler for Reconfigurable Systems

“…Nevertheless, if the execution time of the tasks was variable (for instance, depending on the input data), the scheduler could adopt a solution similar to the one proposed in [23], where the authors suggest creating several graphs for the same task (called scenarios) and then choosing the one that best fits to the current conditions. These scenarios are identified at design-time; hence if it was necessary to apply this technique, the scheduling flow that we propose would continue being valid.…”

A task graph execution manager for reconfigurable multi-tasking systems

“…If the execution time of these tasks heavily depends on the input data we propose to use the same approach as in [7,8]. In this work they include several scenarios for those task graphs with important execution-time variations.…”

“…These threads deal with dynamic events that trigger the execution of one or several task graphs, represented as Direct Acyclic Graphs (DAGs). A similar approach has been successfully applied to deal with complex MPEG-4 and MPEG-2 applications for embedded multi-processor platforms [7,8].…”

Challenging VMs on battery-powered embedded devices

“…There is ongoing research to help with this problem, such as Task Concurrency Management (TCM). TCM Wong et al 2001) is a methodology which balances the resource utilization, performance, and resource manager complexity with a two phase approach (design-time and run-time). TCM helps in finding a good decomposition and balances the resource utilization, performance, and resource manager complexity.…”