Distributed intelligent terminal GPU computing power improvement method, terminal, system and medium

Abstract

The invention discloses a distributed intelligent terminal GPU computing power improving method, terminal, system and medium, and the method comprises the steps: scanning available GPUs on an intelligent terminal and available GPUs on other intelligent terminals near the intelligent terminal, and adding the scanned GPUs into a scheduling queue; and then, according to the task content of the GPU task needing to be executed by the software application on the intelligent terminal and the performance parameters of each GPU in the scheduling queue, the GPU task is distributed to one or more GPUs in the scheduling queue for calculation processing. Therefore, the GPU task processing capability of the local intelligent terminal can be improved by borrowing the GPU hardware computing power of the external intelligent terminal.

Description

technical field [0001] The present invention relates to the technical field of distributed image computing, and in particular to a method, terminal, system and medium for improving the computing power of a distributed smart terminal GPU. Background technique [0002] With the rise of the tide of smart terminals (such as mobile phones, smart cockpits, etc.), GPUs have begun to develop rapidly in mobile devices. However, GPUs in mobile devices are limited by the chip area, energy consumption and cost, and some performance and bandwidth must be sacrificed. Find a balance between cost performance and battery life. However, the pursuit of GPU performance by mobile terminals is constantly expanding. For each generation of mobile chips (such as mobile phone chips), the performance upgrade of GPU is a very critical indicator. [0003] Although the hardware performance of the GPU is constantly improving, the upgrade of the hardware still cannot keep up with the performance requireme...

AI Technical Summary

Problems solved by technology

[0005] 1. The screen projection function is only to passively play the screen of the mobile terminal. The smart terminal system itself is still limited by the insufficient GPU hardware performance, which cannot improve the performance of the smart terminal itself. The system freezes still exist, and the system freezes will also Affects the performance of the screen-casting app function, resulting in a poor screen-casting experience

However, the current screen projection technology can play videos well, but if you run high frame rate applications such as games, there is still obvious lag and poor experience;

[0006] 2. The mobile terminal is considered a relatively private personal item, and it is easy to cause privacy leakage when projecting the screen to the smart terminal;

[0007] 3. There are still differences between the ecology of mobile terminals and the ecology of smart terminals. For example, in the scene where the screen needs to be projected to the central control screen of the car, there will be no feature phones customized for the car factory, so the screen projection solution cannot meet all the scenarios of the smart terminal.

Images