A lightweight class method for system power consumption optimization based on online learning

Abstract

The invention provides a lightweight class method for system power consumption optimization based on online learning. The method comprises the steps of: 1, performing program compiling to generate target codes; 2, starting a monitoring module to monitor hardware events; 3, performing normalization processing of events; 4, building a system power consumption module; 5, designing different optimization modes; 6, designing a value function module; 7, writing the power consumption module, a penalty factor and the value function module into an agent module Agent; 8, designing a software timer and starting the third step and the seventh step regularly; 9, executing a program, the seventh step and the third step and updating the Agent; 10; setting convergence; 11, according to results of the Agent module, going to the second step and starting from the third step until operation is over. Through the steps, temperature, performance and power consumption are comprehensively and synergistically considered and a lightweight class machine learning algorithm is used to search for existing optimization space, so that the effects of low power consumption and reasonable performance are achieved and the problem is solved that working time is influenced because embedded devices are limited by batteries.

Description

technical field [0001] The present invention provides a lightweight method for optimizing system power consumption based on online learning, which relates to the technical field of power consumption optimization for embedded systems, and in particular to a method for combining power consumption optimization for embedded systems with machine learning algorithms. The method is applied in the power consumption optimization and power consumption estimation of the embedded system, and can improve the life cycle and performance of the embedded system. Background technique [0002] Embedded devices have been used more and more in daily life. More embedded terminals and wider online interconnection make embedded system power consumption a problem that designers must face. In addition to energy shortage and environmental protection The current situation of the processor power consumption has attracted more and more attention. Low power consumption has become an important indicator of...

AI Technical Summary

Problems solved by technology

[0006] (1) The common system power consumption optimization is to optimize the design of low power consumption from a single level. Since the power consumption is generated by the software driver through the hardware, the low power consumption design of a single level does not take into account the influence of other levels on power consumption. Factors, so the ability to reduce the overall power consumption of the system is limited;

[0007] (2) For the method of establishing a multi-level system power consumption model, it is necessary to take into account the mutual influence between performance and power consumption. At the same time, the power consumption model is suitable for different application environments, and lacks robustness established on multiple levels. System Power Model

[0008] (3) Common machine learning has the disadvantages of convergence and overfitting, and is not suitable for application in system power consumption optimization, because such algorithms will increase the load on the system, and frequent CPU calculations and I / O calls will increase Dynamic power consumption, it is difficult to reduce the overall power consumption

Images