Device-Cloud Rendering Split for Mobile GPU Load Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The computing power gap between lightweight GPUs in mobile devices and PC-level GPUs, particularly in rendering advanced 3D applications like ray tracing, necessitates a more efficient method to achieve high-quality rendering without overburdening the device's GPU.

Innovation Solution

A device-cloud collaborative rendering method where the mobile device performs basic rendering and a cloud server handles advanced rendering preprocessing, fusing the results to deliver high-quality images with reduced computational load on the device.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If advanced rendering (ray tracing) is performed on mobile device GPU, then rendering quality is improved, but device power consumption and computational load increase significantly

Engineering Contradiction:
Improverendering qualityVSAvoiddevice power consumption
Core Design Contradiction:
Manufacturing precisionVSUse of energy by moving object

Solution Approach 1:

The rendering process is divided into two segments: basic rendering performed on the mobile device and advanced rendering preprocessing performed on the cloud server. This segmentation allows the mobile device to handle only essential rendering tasks while offloading computationally intensive advanced rendering to the cloud, thereby improving rendering quality without significantly increasing device power consumption.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The cloud server acts as an intermediary between the mobile device and the final rendered output. The server receives basic rendering results from the mobile device, performs advanced rendering preprocessing (such as ray tracing), and returns the enhanced results to the mobile device for display. This intermediary approach enables advanced rendering effects without requiring the mobile device to perform all rendering computations locally.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Manufacturing precision

If advanced rendering preprocessing is performed on cloud server, then device-side rendering quality is improved, but cloud-side computing resource consumption increases

Engineering Contradiction:
Improverendering qualityVSAvoidcloud-side computing resource consumption
Core Design Contradiction:
Manufacturing precisionVSLoss of energy

Solution Approach 1:

The cloud server performs only partial advanced rendering preprocessing rather than complete rendering. Specifically, it performs preprocessing tasks such as lighting calculations and shadow maps, while leaving final image composition to the mobile device. This partial action approach provides advanced rendering quality improvements without requiring the cloud server to perform all rendering computations, thereby reducing cloud computing resource consumption.

Inventive Principle:
Principle #16Partial or excessive action

3Productivity

If mobile device performs only basic rendering, then device load is reduced, but rendering quality gap compared to PC-level GPU persists

Engineering Contradiction:
Improvedevice rendering efficiencyVSAvoidrendering quality
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The cloud server performs preliminary advanced rendering preprocessing before the mobile device performs final rendering. This preliminary action includes pre-calculating lighting effects, shadows, and other advanced rendering elements that would be computationally expensive for the mobile device to calculate in real-time. By performing these calculations in advance on the cloud server, the mobile device can achieve higher rendering quality without significantly increasing its real-time computational load.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4691580A1Device-cloud collaboration rendering method and related apparatus
Publication Date: 2026.02.11 HUAWEI TECH CO LTD
  • EP4691580A1 patent drawingFigure 1
  • EP4691580A1 patent drawingFigure 2A
  • EP4691580A1 patent drawingFigure 2BB

AI summary

This application discloses a device-cloud collaborative rendering method. The method includes: An electronic device performs basic rendering on a first device-side scene to obtain first basic rendering data; the electronic device sends a scene identifier of the first device-side scene and a first advanced rendering type to a server; the server stores a first cloud-side scene transformed from the first device-side scene for the first advanced rendering type; the electronic device sends first status data to the server; the server updates the first cloud-side scene based on the first status data, and performs preprocessing of the first advanced rendering type on the first cloud-side scene, to obtain first preprocessing data; the server sends the first preprocessing data to the electronic device; and the electronic device obtains a first image based on the first basic rendering data and the first preprocessing data, and displays the first image. In this way, a computing power requirement for and load of a device-side GPU are reduced, so that an electronic device with a low GPU computing power can also present advanced rendering effect, thereby effectively improving user experience.