Hybrid Test Compression Architecture Using Multiple Codecs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Integrated circuit testing faces challenges in achieving a balance between data compression and coverage, leading to significant test time and cost inefficiencies due to high correlation in test data, which results in reduced quality of manufacturing tests, especially in low-pin count devices.
Innovation Solution
The use of multiple codecs, specifically a high-compression codec followed by a low-compression codec, and optionally a bypass mode, to generate test patterns that recover coverage loss while maintaining efficient compression ratios, thereby reducing test time and data volume.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If high compression ratio techniques are used to reduce test data volume, then data transfer efficiency is improved, but fault coverage is reduced due to higher correlation in test data
Solution Approach 1:
The patent segments the single compression function into two distinct codecs: a high-compression codec for initial compression and a low-compression codec for coverage recovery. This segmentation allows each codec to be optimized for its specific function, resolving the contradiction between compression ratio and fault coverage by applying different compression strategies to different portions of the test data
Solution Approach 2:
The patent changes the compression parameter (compression ratio) by selecting different codecs based on the testing phase and requirements. The system dynamically adjusts the compression parameter between high and low values to balance data volume reduction with fault coverage maintenance, directly addressing the contradiction through parameter optimization
2Reliability
If uncompressed bypass mode is used to recover coverage loss, then fault coverage is improved, but test time increases significantly
Solution Approach 1:
The patent optimizes the compression parameter by introducing a low-compression codec with intermediate compression ratios between high-compression and bypass modes. This parameter adjustment allows coverage recovery without the full test time penalty of uncompressed mode, as the low-compression codec processes data faster than bypass mode while still improving coverage
Solution Approach 2:
The low-compression codec acts as an intermediary between high-compression and bypass modes. It provides a middle-ground solution that recovers coverage loss more efficiently than bypass mode by maintaining some compression while reducing correlation, thus resolving the time-coverage tradeoff
3Device complexity
If the number of scan channels is reduced to lower device complexity, then manufacturing cost is reduced, but compression efficiency must increase which reduces coverage
Solution Approach 1:
The patent segments the compression function into multiple specialized codecs to compensate for reduced scan channel efficiency. By dividing the compression task between high-compression and low-compression codecs, the system maintains adequate fault coverage even with fewer scan channels, resolving the contradiction between device simplicity and testing effectiveness
Solution Approach 2:
The dual-codec architecture serves as an intermediary mechanism that enables reduced scan channel counts while maintaining coverage. The low-compression codec specifically addresses the coverage loss from higher correlation, allowing the system to use fewer scan channels without sacrificing reliability
Data Source
AI summary
This invention uses multiple codecs to efficiently achieve the right balance between compression and coverage for a given design. This application illustrates a simple example using two codecs including a high compression codec and a low compression codec. The test engineer generates a first set of test patterns using the high compression codec. If this high compression results in unacceptable fault coverage loss, the top-up patterns for additional coverage are generated using the low compression codec. The invention may use multiple codecs serially one after the other. The codecs can be of different types or parameters (such as compression ratio, debug tolerance and combinational codec versus sequential codec).