Look Up Table Data Element Promotion for Memory Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern digital signal processors face challenges such as increasing workloads, memory system latency, and unreliable memories and registers, which affect the efficiency of look up table operations.
Innovation Solution
The implementation of special purpose look up table read operations that allow for data element promotion to larger sizes, enabling extension bits for recalled data elements and treating them as unsigned or signed integers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If two tables are stored and employed to provide data in two precisions, then the data processing capability is improved, but the memory space is wasted
Solution Approach 1:
The patent merges multiple tables with different data precisions into a single table structure. The table stores data elements in a base precision (e.g., 8-bit), and the processor can retrieve and promote these elements to higher precisions (16-bit, 32-bit, 64-bit) as needed. This eliminates the need to maintain separate tables for different precisions, reducing memory usage while preserving the ability to process data in multiple precisions.
Solution Approach 2:
The patent changes the parameter of data precision dynamically. Instead of storing data in fixed precision formats for different operations, the system stores data in a base precision and changes the precision parameter during processing through promotion operations. The processor can promote data elements from the table to larger sizes (2×, 4×, or 8× the base size) depending on the computational requirements, allowing flexible precision adjustment without requiring multiple dedicated tables.
2Quantity of substance
If table values are stored at discrete intervals, then the memory usage is reduced, but the filtering algorithm cannot handle inputs between the stored values
Solution Approach 1:
The patent applies parameter changes by promoting data elements from a base precision to higher precisions. When table values are stored at discrete intervals in a base precision, the system can promote these values to larger precisions (2×, 4×, or 8×) to provide finer granularity for filtering operations. This allows the same table structure to serve both memory-efficient discrete lookup and high-precision filtering requirements.
Data Source
AI summary
Disclosed embodiments relate to look up table operations implemented in a digital data processor. A look up table read instruction recalls data elements of a specified data size from table(s) and stores recalled data elements in successive slots in a destination register. Disclosed embodiments promote data elements to a larger size with selected sign or zero extension. A source operand register stores vector offsets from a table start address. A destination operand stores the results of the look up table read. The look up table instruction implies a base address register and a configuration register. The base address register stores a table base address. The configuration register sets various look up table read operation parameters.