Frame Buffer Bit-Splitting for Lower Streaming Data Bandwidth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for storing and retrieving streaming data, such as video data, are inefficient as they fail to distinguish between frames with different data requirements, leading to suboptimal memory bandwidth utilization and performance degradation in systems like motion vector processing and LCD display response-time compensation.
Innovation Solution
The approach involves separating incoming data into most significant bits (MSB) and least significant bits (LSB), packing them into separate words, and storing them in distinct areas of a frame buffer, allowing for independent access and recombination based on the application's needs, thereby reducing memory bandwidth requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If all frames are stored and accessed using the same data width (P bits), then the system maintains simplicity in storage architecture, but memory bandwidth utilization becomes inefficient when different frames require different amounts of data
Solution Approach 1:
The patent divides each frame's data into two separate segments: MSB data and LSB data. These segments are stored in separate memory areas (first memory area for MSB, second memory area for LSB). This segmentation allows the system to selectively access only the required data segment for each frame, improving memory bandwidth utilization without requiring a completely complex new architecture
Solution Approach 2:
The patent implements dynamic data access by determining whether LSB data is needed for each frame before accessing memory. The system dynamically adjusts the access pattern: for some frames it accesses only MSB data (M bits), while for other frames it accesses both MSB and LSB data (P bits). This dynamic approach optimizes memory bandwidth utilization based on actual frame requirements
2Productivity
If the system accesses all P bits of data for every frame, then the retrieval process remains simple and uniform, but memory bandwidth is wasted when only M bits (M < P) are actually needed
Solution Approach 1:
The patent extracts and separates the LSB data from the MSB data, storing them in different memory areas. This extraction allows the system to take out only the necessary data (either just MSB or both MSB and LSB) for each frame retrieval operation, preventing the waste of memory bandwidth that would occur if all P bits were accessed uniformly for every frame
Solution Approach 2:
The patent applies local quality by assigning different data precision requirements to different frames based on their specific needs. Instead of uniformly applying P-bit precision to all frames, the system uses M-bit precision for frames where LSB data is not needed and P-bit precision for frames where it is needed, optimizing memory bandwidth utilization by matching data quality to actual requirements
Data Source
AI summary
Various technologies for efficiently storing and retrieving streaming data are described. Bits of data (e.g., P bits of data) are received and separated into most significant bits (MSB) of data and least significant bits (LSB) of data. Further, the MSB of data and the LSB of data are respectively packed into a first word and a second word. The first word is stored in a first area of a frame buffer and the second word is stored in a second area of the frame buffer. As a result, data is managed in a more efficient way to reduce memory bandwidth requirement.