Reducing scan-shift power through scan partitioning and test vector reordering
Tiebin Wu, Li Zhou, Hengzhu Liu
Abstract
Tiebin Wu, Li Zhou, Hengzhu Liu
Abstract
Excessive test power dissipation results in over-testing, IR-drop, yield loss and even heat damage to the circuit under test (CUT). An efficient scan-shift power reduction scheme based on scan chain partitioning and test vector reordering is presented in this paper. After partitioning the scan chains into several segments equally, a heuristic ant colony optimization (ACO) algorithm is introduced to reorder the test vectors to minimize the clashes between the previous test response and current test vector, which leads to scan-shift power reduction further. Experimental results show that the proposal can achieve 3.48% scan-shift power reduction on average with the help of ACO test vectors reordering after scan partitioning. Furthermore, the proposed scan-shift power reduction technique can be acceptable for any scan-based testing architecture without affecting test application time, test fault coverage, performance and/or routing cost of the CUT.
OpenAlex reports 10 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Excessive test power dissipation results in over-testing, IR-drop, yield loss and even heat damage to the circuit under test (CUT). An efficient scan-shift power reduction scheme based on scan chain partitioning and test vector reordering is presented in this paper. After partitioning the scan chains into several segments equally, a heuristic ant colony optimization (ACO) algorithm is introduced to reorder the test vectors to minimize the clashes between the previous test response and current test vector, which leads to scan-shift power reduction further. Experimental results show that the proposal can achieve 3.48% scan-shift power reduction on average with the help of ACO test vectors reordering after scan partitioning. Furthermore, the proposed scan-shift power reduction technique can be acceptable for any scan-based testing architecture without affecting test application time, test fault coverage, performance and/or routing cost of the CUT.
Key concepts: Scan chain, Test vector, Test compression, Automatic test pattern generation, Reduction (mathematics), Computer science, Fault coverage, Power (physics)