Answers are grounded in this paper and cite their sources.
For current highly-integrated and memory-dominant system-on-a-chips (SoCs), especially for graphics and networking SoCs, the test/repair area overhead of embedded SRAMs (e-SRAMs) is a big concern. This paper presents various approaches to tackle this problem from a practical point of view. Without sacrificing at-speed testability, diagnosis capability and repairability, the proposed approaches consider partly sharing wrapper for identical memories, sharing memory BIST controllers for e-SRAMs embedded in different functional blocks, test responses compression for wide memories, and various repair strategies for e-SRAMs with different configurations. By combining the above approaches, the test/repair area overhead for e-SRAMs can be significantly reduced. For example, for one benchmark SoC used in our experiments, it can be reduced as much as 10% of the entire memory array
No takes yet. Share an insight, caveat, or question.
Wang et al. (2006) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: