Scenario-based design space exploration of MPSoCs

Abstract

Abstract-Early design space exploration (DSE) is a key ingredient in system-level design of MPSoC-based embedded systems. The state of the art in this field typically still explores systems under a single, fixed application workload. In reality, however, the applications are concurrently executing and contending for system resources in such systems. As a result, the intensity and nature of application demands can change dramatically over time. This paper therefore introduces the concept of workload scenarios in the DSE process, capturing dynamic behavior both within and between applications. More specifically, we present and evaluate a novel, scenario-based DSE approach based on a coevolutionary genetic algorithm.

A trace-based scenario database for high-level simulation of multimedia MP-SoCs

“…Specifically, we have compared the storage size of the database to the storage size of the plain traces. As mentioned before, actual deployment of the database in real system-level MPSoC simulations and DSE is beyond the scope of this paper, but the interested reader is referred to [5], [6] for a study on scenario-aware DSE using the scenario database.…”

“…Such encodings can be easily used in, for example, design space exploration experiments based on genetic algorithms [5], [6] (which is beyond the scope of this paper). A complete application scenario can be encoded as a concatenation of the inter-application scenario and the table entries of the intra-application scenarios for each of the applications.…”

“…This database can be used as input for Sesame for on-demand generation of multi-application workloads. For example, we have deployed the database for on-demand workload generation in our recently-proposed, scenario-based DSE [5], [6] which simultaneously searches for optimal MP-SoC design instances and for representative workload scenarios to evaluate these design instances.…”

“…Actual deployment of the database in real system-level MP-SoC DSE experiments is, however, beyond the scope of this paper. The interested reader is referred to [5], [6], in which the scenario database has been used for a new scenario-based DSE method which simultaneously searches for optimal MP-SoC design instances and for a representative set of workload scenarios to evaluate these design instances.…”

A hierarchical run-time adaptive resource allocation framework for large-scale MPSoC systems

“…At design time, a performance-optimized task mapping (and, if needed, also a poweroptimized mapping) for each execution mode of each application in isolation is determined by using state-of-the-art scenario-aware Design Space Exploration (DSE) techniques [26,40]. Note that the target architecture in this optimisation process is the tile architecture inside the target MPSoC system as mentioned above.…”