Foveated Rendering for GPU Power Reduction in HMDs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current graphics processing units (GPUs) consume excessive power by applying uniform brightness and contrast to all display regions, leading to inefficiencies in power management and reduced battery life in devices like Head-Mounted Displays (HMDs) used for VR/AR/MR applications.
Innovation Solution
Implement low power foveated rendering techniques, where the fovea region is rendered with high brightness and contrast, while surrounding areas have lower brightness and contrast, and variable compression rates are applied based on pixel location, reducing power consumption by using different brightness and contrast settings across the display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If uniform brightness and contrast are applied to all display regions, then display quality is maintained across the entire screen, but power consumption increases
Solution Approach 1:
The patent applies different brightness and contrast settings to different regions of the display based on where the user is looking. The fovea region (center of gaze) receives high brightness and contrast for sharp visual acuity, while peripheral regions receive lower brightness and contrast. This local differentiation maintains perceived display quality while reducing overall power consumption, as the human visual system is less sensitive to quality variations in peripheral vision.
2Measurement precision
If high resolution rendering is applied to the entire display, then visual quality is maximized, but GPU power consumption increases
Solution Approach 1:
The patent segments the display area into multiple regions based on user gaze direction, typically dividing the screen into a fovea region (high priority) and peripheral regions (low priority). Different rendering resolutions are assigned to each segment: high resolution for the fovea region where the user is looking, and lower resolution for peripheral regions. This segmentation allows the GPU to focus computational resources on the most visually important areas, maintaining perceived quality while reducing overall power consumption.
Solution Approach 2:
The patent applies local quality differentiation by rendering only the fovea region (area of user gaze) at high resolution, while rendering peripheral regions at lower resolution. This approach leverages the human visual system's characteristic of having high acuity only in the central foveal region and reduced acuity in peripheral regions. The GPU processes fewer pixels at high quality levels, significantly reducing computational load and power consumption while maintaining acceptable visual quality in the user's actual field of view.
3Illumination intensity
If full brightness is used across the entire display, then visibility is maximized, but battery life decreases
Solution Approach 1:
The patent implements local quality control by setting high brightness only in the fovea region (center of user gaze) where visibility is most critical, and reducing brightness in peripheral regions where the human eye is less sensitive. This creates a brightness gradient that maintains adequate visibility in the area the user is actually looking at, while conserving battery power by dimming areas that contribute less to perceived visibility. The result is extended battery life without significantly compromising the user's viewing experience.
Data Source
AI summary
Methods and apparatus relating to techniques for provision of low power foveated rendering to save power on GPU (Graphics Processing Unit) and/or display are described. In various embodiment, brightness/contrast, color intensity, and/or compression ratio applied to pixels in a fovea region are different than those applied in regions surrounding the fovea region. Other embodiments are also disclosed and claimed.