Thermal-sensing and low-energy consumption task scheduling method oriented to heterogeneous MPSoC system

Abstract

The invention discloses a thermal-sensing and low-energy consumption task scheduling method oriented to a heterogeneous MPSoC system. The method comprises the following steps: converting a heterogeneous multi-processor supporting multiple frequencies into a virtual multi-processor model only supporting a fixed frequency; distributing the task with the dependency relation to the virtual processor by using a heuristic algorithm; computing peak temperature of the system under the current task distribution; computing the accomplishment time of the task under the current task distribution; verifying this task distribution satisfies a constraint condition of the time and the task dependency relation; and updating a undistributed task queue, the task queue distributed to each virtual processor, and the task execution time of each virtual processor. Under the constraint of the task expiration deadline, the task priority and the chip peak temperature, the system energy consumption is optimizedthrough the thermal-sensing energy-saving scheduling algorithm; the algorithm determines the task distributed to each processor and the working efficiency of each task, thereby effectively computing the low energy-consumption task scheduling solution.

Description

technical field [0001] The invention relates to a task scheduling method for an embedded system, in particular to a task scheduling scheme that comprehensively considers task deadline, task priority and chip peak temperature, in particular to a heterogeneous MPSoC system supporting multiple frequencies A task scheduling method to reduce processor power consumption. Background technique [0002] As the demand for high performance computing and hardware performance continues to increase, so does the power consumption of multiprocessors, which affects the lifespan of portable devices with limited battery capacity. At the same time, the increase in performance has led to a sharp increase in the integration of transistors in multiprocessors, which increases the power density of the chip and thus increases the chip temperature. This high temperature can accelerate the wear of the multiprocessor through electromigration, dielectric breakdown, thermal cycling, etc., thereby reducin...

AI Technical Summary

Problems solved by technology

This high temperature will accelerate the wear and tear of the multiprocessor through electromigration, dielectric breakdown, thermal cycling, etc., thereby reducing the reliability of the system

And as the temperature rises, so does the leakage power, which in turn increases the die temperature and leads to increased packaging and cooling costs

Therefore, the increase in energy consumption and chip temperature may lead to serious economic, ecological and technical problems, and minimizing energy consumption under temperature constraints is an urgent research problem in the design of sustainable computing systems

Images