Residual Data Writing Using Parallel Buffer Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for writing residual data in video encoding and decoding are inefficient, leading to long writing times and low access efficiency of residual data, which significantly affects the encoding efficiency.
Innovation Solution
A residual data writing method and apparatus that determines a residual data reading mode, configures buffer groups and buffers, and writes residual data into corresponding buffers using a unique buffer index and buffer address, allowing simultaneous writing of residual data for different coordinates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If residual data is written sequentially in row scanning sequence, then the data can be stored in SRAM, but the writing time is long and access efficiency is low
Solution Approach 1:
The patent divides the SRAM into multiple banks (e.g., Bank0 and Bank1) and further segments each bank into multiple buffers. This segmentation allows parallel writing operations across different banks, reducing the total writing time while maintaining data storage reliability.
Solution Approach 2:
The patent introduces a new dimension of parallelism by organizing buffers in a two-dimensional structure (bank × buffer index) and using multiple data interfaces to simultaneously write residual data from multiple motion vectors into different buffers within the same clock cycle, thereby significantly reducing writing time.
2Quantity of substance
If residual data for multiple motion vectors is buffered sequentially, then all data can be stored, but the encoder remains idle for long periods, affecting encoding efficiency
Solution Approach 1:
The patent performs preliminary organization of residual data by buffer groups and buffer indices before writing, allowing the encoder to prepare data access patterns in advance. This enables continuous data flow to multiple buffers simultaneously, eliminating idle periods and improving encoding efficiency while maintaining full data capacity.
Solution Approach 2:
By configuring multiple buffer groups with unique buffer indices and using multiple data interfaces, the patent enables continuous writing operations without idle cycles. The encoder can continuously feed residual data from multiple motion vectors into different buffers in parallel, maintaining productive action throughout the buffering process.
3Quantity of substance
If multiple pieces of residual data are written to the same buffer address, then data can be stored, but time redundancy occurs and access efficiency decreases
Solution Approach 1:
The patent assigns different buffer indices to different motion vectors within the same bank, creating local differentiation in the buffer addressing scheme. This allows each motion vector's residual data to be directed to a specific buffer location, eliminating time redundancy and enabling efficient parallel access while maintaining full data storage capacity.
Data Source
AI summary
Disclosed are a residual data writing method and apparatus, a computer device, and a storage medium. The method includes: determining a residual data reading mode comprising a reading sequence of each piece of residual data, and the quantity of residual data read simultaneously at each time; determining the number of buffer groups and the number of buffers in each buffer group according to the residual data reading mode, wherein each buffer is provided with a unique buffer index, and each buffer comprises a plurality of buffer addresses; acquiring a buffer index and a buffer address corresponding to to-be-written residual data; and writing the to-be-written residual data into a corresponding buffer according to the buffer index and the buffer address.