Dynamic Tiling and Dirty Tile Marking for VR Latency

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current virtual reality systems face challenges in optimizing time/space warp techniques for efficient rendering, particularly in maintaining image fidelity and reducing latency, due to limitations in tile-based rendering architectures that result in noticeable edge losses and increased latency.

Innovation Solution

The implementation of dynamic tiling and dirty tile marking techniques, which subdivide images into core and peripheral tiles, allowing for larger image rendering and efficient transmission, and optimizing time/space warp by predicting head motion to concentrate rendering efforts on the predicted region of interest.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If traditional tile-based rendering is used for virtual reality, then memory bandwidth is reduced, but image fidelity deteriorates with noticeable edge losses

Engineering Contradiction:
Improvememory bandwidthVSAvoidimage fidelity
Core Design Contradiction:
Use of energy by moving objectVSManufacturing precision

Solution Approach 1:

The patent divides the frame buffer into multiple tiles, where each tile is a manageable unit of rendering. By segmenting the rendering area into tiles, the system can process and transmit image data in smaller chunks, reducing memory bandwidth requirements while maintaining overall image quality through selective rendering of relevant tiles.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements selective rendering where only tiles that contain relevant visual information (dirty tiles) are rendered with high fidelity, while other tiles may use lower quality or be skipped entirely. This local quality approach maintains image fidelity in critical areas while reducing overall memory bandwidth consumption.

Inventive Principle:
Principle #3Local quality

2Use of energy by moving object

If traditional tile-based rendering is used for virtual reality, then memory bandwidth is reduced, but rendering latency increases

Engineering Contradiction:
Improvememory bandwidthVSAvoidrendering latency
Core Design Contradiction:
Use of energy by moving objectVSLoss of time

Solution Approach 1:

The patent predicts future head motion and pre-identifies tiles that will be needed in the near future. By performing preliminary identification of dirty tiles based on predicted head movement, the system can prepare and transmit these tiles in advance, reducing rendering latency while maintaining efficient memory bandwidth usage.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements a feedback mechanism where head tracking data is continuously monitored and used to update predictions about which tiles will be needed. This feedback loop allows the system to adaptively adjust tile rendering priorities and timing, optimizing both latency and memory bandwidth utilization.

Inventive Principle:
Principle #23Feedback

3Productivity

If head motion prediction is implemented, then rendering efficiency is improved, but computational complexity increases

Engineering Contradiction:
Improverendering efficiencyVSAvoidcomputational complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent implements head motion prediction with a limited prediction horizon, only forecasting head movement for a short time window into the future. This partial action approach provides sufficient rendering efficiency improvement without requiring complex long-term prediction algorithms, thus balancing productivity gain against computational complexity.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS10504272B2Apparatus and method for optimizing time/space warp for virtual reality using dynamic tiling and dirty tile marking
Publication Date: 2019.12.10 INTEL CORP
  • US10504272B2 patent drawing
  • US10504272B2 patent drawing
  • US10504272B2 patent drawing

AI summary

An apparatus and method for efficient rendering and transmission of video content in a virtual reality system. For example, one embodiment of an apparatus comprises: a first frame buffer to store image frames to be transmitted to a virtual reality apparatus; tile-based image rendering circuitry and/or logic to concurrently render multiple tiles of a first image frame, wherein tiles are categorized core tiles or a peripheral tiles; tile-based transmission circuitry and/or logic to transmit a first core tile rendered to the virtual reality apparatus before one or more remaining tiles have been rendered, the tile-based transmission circuitry and/or logic to continue to transmit one or more subsequently rendered core tiles following the first tile until all core tiles have been transmitted; and the tile-based transmission circuitry and/or logic to identify a subset of peripheral tiles to be transmitted based on first coordinate/offset data and to responsively transmit the subset of peripheral tiles.