A multi-core SoC software mapping method based on multi-objective optimization

Abstract

The present invention relates to a kind of multi-core SoC software mapping method based on multi-objective optimization, which comprises the following steps: (1), optimize the target of multi-core SoC software mapping; (11), determine the decision variable in the multi-core SoC software mapping; (12) 1. Determine the constraints of multi-core SoC software mapping; (13), determine the target total function of multi-core SoC software mapping (14) establish the mapping overhead between the two based on tasks and processors; (15) establish the relationship between communication events and processors (2) The multi-objective evolution steps of multi-core SoC software mapping are as follows: (21) Coding; (22) Initial chromosome population generation; (23) Constraint violation checking; (24) Fitness calculation; (25) selection; (26) mutation; (27) hybridization; (28) repeating steps (23) to (27) until the predetermined evolution algebra is completed.

Description

technical field [0001] The invention belongs to the field of computer communication and relates to a multi-core SoC software mapping method based on multi-objective optimization. Background technique [0002] The emergence of deep submicron technology has made it possible to integrate the entire computer system into a chip, that is, a system-on-a-chip. System chip SoC (System-on-a-Chip), also known as system-level integrated circuit SLI, refers to the integration of digital and analog hybrid circuits on a single silicon chip, including embedded cores, memories, special function modules, and I / O interfaces. Modules and other modules with multiple functions, that is to say, existing standard integrated circuits, application-specific integrated circuits, memories, logic circuits, analog circuits, PLDs, etc. can be integrated into one chip. [0003] Multi-core SoC technology has become the main computing architecture of portable computing devices. It provides strong computing p...

AI Technical Summary

Problems solved by technology

[0004] With the development of aerospace and national defense, computer systems are facing some problems: 1. The demand for computing power is increasing day by day; 2. The energy consumption of computing is too large

However, due to the complexity of the multi-core SoC architecture and the evolution of the application itself, it is usually very difficult to obtain an optimal or suboptimal mapping scheme

Images