Symbol Boundary Extraction for High-Throughput Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing decoding devices struggle with high latency and inefficiency in decoding large amounts of compressed data due to the need to sequentially decode symbols one by one, which limits throughput.
Innovation Solution
The decoding device employs extraction circuitry to determine the starting location of each symbol based on boundary location information, allowing the dictionary-based decoding circuitry to operate independently, thus eliminating the need for sequential decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If sequential decoding of symbols one by one is used, then decoding accuracy is maintained, but decoding speed and throughput deteriorate
Solution Approach 1:
The patent pre-calculates and stores boundary location information indicating the starting positions of symbols in the compressed data before actual decoding. This preliminary action allows the decoding process to jump directly to symbol boundaries without sequential processing, significantly improving both decoding speed and throughput while maintaining accuracy.
2Loss of time
If sequential decoding process is used, then decoding accuracy is ensured, but latency increases
Solution Approach 1:
Boundary location information is pre-computed and stored, indicating the starting positions of all symbols in advance. This eliminates the time-consuming sequential search for symbol boundaries during decoding, thereby reducing latency and improving decoding efficiency simultaneously.
3Productivity
If extraction circuitry with boundary location information is implemented, then decoding throughput is improved, but device complexity increases
Solution Approach 1:
The decoding device is segmented into distinct functional modules: extraction circuitry that retrieves boundary location information, and dictionary-based decoding circuitry that performs actual decoding. This segmentation allows parallel operation of modules, improving throughput while keeping each module's complexity manageable through clear functional separation.
Data Source
AI summary
According to one embodiment, extraction circuitry of a decoding device calculates a first code length of a first symbol and a second code length of a second symbol. The extraction circuitry determines first context information corresponding to a third symbol, and second context information corresponding to a fourth symbol, and stores the first context information and first boundary location information, and the second context information and second boundary location information. The extraction circuitry acquires third context information and third boundary location information corresponding to a fifth symbol following the third symbol, and fourth context information and fourth boundary location information corresponding to a sixth symbol following the fourth symbol.