M3-SCoPE: Performance Modeling of Multi-Processor Embedded Systems for Fast Design Space Exploration - Multi-objective Design Space Exploration of Multiprocessor SoC Architectures - page 19

Chapter 2
M3-SCoPE: Performance Modeling
of Multi-Processor Embedded Systems
for Fast Design Space Exploration

Hector Posadas, Sara Real, and Eugenio Villar

Abstract Design Space Exploration for complex, multi-processor embedded sys-
tems demands new modeling, simulation, performance estimation tools and design
methodologies. Recently approved as IEEE 1666 standard, SystemC has proven to
be a powerful language for system modeling and simulation. In this chapter, M3-
SCoPE, a SystemC framework for platform modeling, SW source-code behavioral
simulation and performance estimation of multi-processor embedded systems is pre-
sented. Using M3-SCoPE, the application SW running on the different processors
of the platform can be simulated efficiently in close interaction with the rest of the
platform components. In this way, fast and sufficiently accurate performance met-
rics of the system are obtained. These metrics are then delivered to the DSE tools to
evaluate the quality of the different configurations in order to select the best ones.

2.1 Introduction

System exploration with an optimum trade-off between performance and cost re-
quires analyzing the performance of a large number of system configurations with a
wide set of parameters, such as number and type of processors, memory architecture
and sizing, mapping of SW tasks and suitability of communication infrastructure.
System simulation is a key design task widely used for design verification and
evaluation. The main role of system simulation in embedded system design is to
ensure the functional correctness of the design at the different abstraction levels. In
Design Space Exploration (DSE), system simulation is used for performance analy-
sis, providing to the exploration tool the required metrics such as delays, throughput,
utilization rates, bandwidths, etc. Power consumption is becoming an additional,
increasingly important metric to be estimated.

H. Posadas ( )
University of Cantabria, Santander, Spain
e-mail: [e-mail]