Guaranteed 10-Bit Pixel Compression for Lower Memory Bandwidth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of efficiently compressing image data in GPUs and CPUs is exacerbated by limited memory bandwidth and space, with existing methods often failing to achieve consistent compression ratios and causing inefficiencies in memory allocation and access.
Innovation Solution
A method is described for converting 10-bit pixel data to 8-bit pixel data, using truncation and error value generation, and employing a combination of lossless and lossy compression techniques to guarantee a predefined compression threshold, ensuring reduced memory bandwidth and space usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If data compression is applied to reduce memory bandwidth consumption, then power consumption is reduced, but memory allocation and access efficiency deteriorate
Solution Approach 1:
The patent segments the 10-bit pixel data into multiple components (e.g., 8-bit color data and 2-bit error data) and stores them in separate memory locations or data structures. This segmentation allows the system to use optimized memory allocation strategies for different data types while maintaining compression benefits, thereby resolving the contradiction between power reduction through compression and memory access efficiency.
Solution Approach 2:
The patent changes the parameter representation by converting 10-bit data into a compressed format with different bit allocations (e.g., 8-bit + 2-bit error). This parameter transformation enables more efficient memory addressing and access patterns while maintaining the compressed size, thus improving both power efficiency and memory access characteristics simultaneously.
2Measurement precision
If higher quality rendering algorithms are used, then rendering quality is improved, but memory bandwidth consumption increases
Solution Approach 1:
The patent creates a composite data structure combining 8-bit color information with 2-bit error data, forming a hybrid representation that preserves high rendering quality while reducing overall data volume. This composite approach allows quality-preserving algorithms to operate on compressed data, simultaneously achieving high rendering quality and reduced memory bandwidth consumption.
3Measurement precision
If 10-bit pixel data is stored to maintain data quality, then data quality is preserved, but memory space and bandwidth requirements increase
Solution Approach 1:
The patent extracts the most significant 8 bits of the 10-bit pixel data for primary storage, and separates the remaining 2 bits as error data stored in alternative formats or locations. This extraction strategy preserves essential data quality in the main 8-bit channel while minimizing the memory footprint of the full 10-bit representation, thus resolving the contradiction between data quality preservation and memory space reduction.
Data Source
AI summary
A method of converting 10-bit pixel data (e.g. 10:10:10:2 data) into 8-bit pixel data involves converting the 10-bit values to 7-bits or 8-bits and generating error values for each of the converted values. Two of the 8-bit output channels comprise a combination of a converted 7-bit value and one of the bits from the fourth input channel. A third 8-bit output channel comprises the converted 8-bit value and the fourth 8-bit output channel comprises the error values. In various examples, the bits of the error values may be interleaved when they are packed into the fourth output channel.