| Research Support Tool |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
A scheme for multiple on-chip signature checking for embedded SRAMS
Title:
A scheme for multiple on-chip signature checking for embedded SRAMS
Archive:
Eprints@IIT Delhi
Author(s):
Abdulla, M F
Ravikumar, C P
Anshul Kumar
Ravikumar, C P
Anshul Kumar
Date:
2005-08-25
Abstract:
Embedded read/write memories are integral parts of many VLSI chips designed for specific applications in the areas of computer communications, multimedia, and digital signal processing. Testing an embedded memory poses a challenge
to a system test engineer, due to its limited controllability and observability. In this paper, we propose a pseudorandom
built-in self test (BIST) scheme to solve this problem. Our technique is based on a test architecture known as multiple on-line signature checking (MOSC) which overs a very low aliasing probability and a high degree of confidence in testing. While the MOSC scheme is suficiently general and applicable to any digital circuit, it can especially be optimized for circuits with embedded memories. We present interesting test scheduling algorithms that reduce the
overhead of testing. On several industry-standard benchmark circuits, we report up to 35% savings in test area overhead.
to a system test engineer, due to its limited controllability and observability. In this paper, we propose a pseudorandom
built-in self test (BIST) scheme to solve this problem. Our technique is based on a test architecture known as multiple on-line signature checking (MOSC) which overs a very low aliasing probability and a high degree of confidence in testing. While the MOSC scheme is suficiently general and applicable to any digital circuit, it can especially be optimized for circuits with embedded memories. We present interesting test scheduling algorithms that reduce the
overhead of testing. On several industry-standard benchmark circuits, we report up to 35% savings in test area overhead.
Index terms:
Discipline(s):
Embedded memories
Subject(s):
Built-in self test; Memory-testing; Multiple signature comparison; Aliasing
Method/Approach:
Coverage:
Publisher:
Contributors:
Source:
Language:
en
Relation:
Type:
Article
Format:
807031 bytes
application/pdf
Copyright Information: