Cost-Quality Trade-offs of Approximate Memory Repair Mechanisms for Image Data

Qianqian Fan, Sachin Sapatnekar, David Lilja
University of Minnesota


Abstract

The traditional approach for increasing yield in large memory arrays has been to eliminate all hard errors using repair mechanisms. However, the cost of these mechanisms can become prohibitive for cheaper memories, which have higher error rates. Instead of completely repairing faulty cells, this paper introduces new approximate memory repair mechanisms that only partially repair both CMOS DRAMs and STT-MRAMs. By combining redundant repair with unequal protection, such as skewing the limited spare elements available for repairing faults towards the k most significant bits, and a hybrid bit-shuffling and redundant repair scheme, the new mechanisms maintain excellent output quality while substantially reducing the cost of the repair mechanism, particularly for increasingly important cluster faults.