Adaptive Lossy Delta Compression for Framebuffer Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current graphics rendering systems face high memory bandwidth and power consumption due to frequent framebuffer operations, with existing compression techniques being complex and potentially reducing image quality or being inefficient in tile-based systems.
Innovation Solution
An adaptive lossy delta-based compression method is introduced, utilizing a Sampling Unit (SU) to capture color characteristics and an update rate, and a Compression Unit (CU) to apply lossless or lossy compression based on tile update rates, with double on-chip tile buffers for efficient data handling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If lossless compression is applied to framebuffer tiles, then memory bandwidth and power consumption are reduced, but image quality may be compromised
Solution Approach 1:
The compression unit dynamically adapts the compression method (lossless or lossy) and compression ratio based on the update rate of each tile. Tiles with low update rates use more aggressive compression, while tiles with high update rates use lossless compression, optimizing the balance between power consumption and image quality in real-time
Solution Approach 2:
The system changes the compression parameter (error rate threshold) based on tile characteristics and update frequency. By adjusting the allowable error rate parameter, the system can switch between lossless and lossy compression modes, reducing power consumption while maintaining acceptable image quality
2Loss of energy
If complex compression techniques are applied to framebuffer data, then memory bandwidth is reduced, but device complexity increases
Solution Approach 1:
The framebuffer is divided into multiple tiles, and each tile is compressed independently based on its own update rate characteristics. This segmentation allows the use of simpler compression algorithms on each tile while achieving overall bandwidth reduction, avoiding the need for a single complex compression system
Solution Approach 2:
The compression unit automatically determines the appropriate compression method and parameters for each tile based on its update rate, without requiring external control or complex centralized decision-making. Each tile effectively compresses itself based on its characteristics, reducing overall system complexity
3Loss of energy
If lossy compression is applied to reduce memory bandwidth, then power consumption is reduced, but data loss increases
Solution Approach 1:
Different compression quality levels are applied to different tiles based on their local characteristics (update rates). Tiles that are updated frequently maintain high quality (lossless compression), while tiles updated rarely can tolerate higher loss, optimizing the balance between power consumption and data preservation locally for each tile
Solution Approach 2:
The compression quality is dynamically adjusted based on the update rate of each tile. When a tile's update rate changes, the compression unit adapts the error rate threshold accordingly, ensuring that data loss remains acceptable while maximizing power savings under varying operational conditions
Data Source
AI summary
The device and method described in this application relate generally to graphics processing systems utilizing the tile based rendering technique and more specifically relate to the processing of the framebuffer data in graphics processing applications. The present invention discloses techniques to reduce the bandwidth needed to access the color data stored in the framebuffer. A method for adaptive lossy delta based compression of color data is disclosed. The error rate, that is the amount of color data lost during the lossy compression process, is controlled by various parameters of the rendered tiles produced by the graphics processing system. The compression process is driven by a dedicated unit which enables informed compression decisions with controllable error rate so as the output color data can be reliably decompressed to produce the original color data with minimal or no errors.