Alpha-Based Pixel Block Compression for Guaranteed Memory Savings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data compression methods for GPUs and CPUs face challenges in reducing memory bandwidth and storage space efficiently, especially when dealing with image data, due to varying compression ratios and the need for random access, which can lead to increased power consumption and memory inefficiencies.
Innovation Solution
A lossy data compression method that divides blocks of pixels into sub-blocks, analyzes alpha channel values to select from a set of candidate compression modes, including fixed and variable alpha modes, to ensure a guaranteed compression threshold is met, thereby reducing memory bandwidth and storage space while enabling random access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data compression is applied to reduce memory bandwidth and storage space, then memory efficiency improves, but compression ratio variability and random access requirements can increase power consumption
Solution Approach 1:
The patent divides image data into multiple blocks, with each block containing multiple pixels. This segmentation allows independent compression analysis of each block, enabling selective application of compression techniques that balance bandwidth reduction with power consumption constraints while maintaining random access capability.
Solution Approach 2:
The patent analyzes alpha channel values as a key parameter to determine compression strategy. By changing compression parameters based on alpha value characteristics (transparent vs. opaque regions), the system achieves consistent compression ratios that reduce memory bandwidth requirements without excessive power consumption increases.
2Quantity of substance
If lossy compression is applied to reduce data size, then storage space decreases, but image quality may deteriorate
Solution Approach 1:
The patent applies different compression strategies to different regions within image blocks based on local characteristics, particularly alpha channel values. Transparent regions use one compression approach while opaque regions use another, preserving image quality in critical areas while achieving compression in less sensitive areas.
Solution Approach 2:
The patent selectively applies lossy compression only where acceptable, using alpha channel analysis to identify regions where compression can be applied without perceptible quality loss. This partial application of compression maintains overall image quality while achieving storage space reduction.
3Stability of the object's composition
If compression modes are selected based on alpha channel analysis, then compression consistency improves, but processing complexity increases
Solution Approach 1:
The patent performs preliminary analysis of alpha channel values for each block before applying compression. This preliminary action identifies the appropriate compression mode in advance, ensuring consistent compression ratios are achieved without requiring complex real-time decision-making during the compression process itself.
Solution Approach 2:
The patent uses alpha channel value parameters as the basis for selecting compression modes. By changing compression parameters based on these pre-analyzed values, the system achieves consistent compression ratios while keeping processing complexity manageable through straightforward parameter-based decision logic.
Data Source
AI summary
Lossy methods and hardware for compressing data and the corresponding decompression methods and hardware are described. The lossy compression method comprises dividing a block of pixels into a number of sub-blocks and then analysing, for each sub-block, and selecting one of a candidate set of lossy compression modes. The analysis may, for example, be based on the alpha values for the pixels in the sub-block. In various examples, the candidate set of lossy compression modes comprises at least one mode that uses a fixed alpha channel value for all pixels in the sub-block and one or more modes that encode a variable alpha channel value.