Memory yield and repair rate improvement scheme using built in self repair techniques

Abstract

Memory is an important part of every computing system. In SOC, 90 to 92% of the total chip area is covered by embedded memories (ITRS 2009) and that means memory density is higher than the logic density. Therefore testing and diagnosis of memories are important issues in the SOCs. Yield of memory is affected by the faults present in memory which also affects the yield of SOC. Built in self-repair techniques are used to repair the embedded memories. Built in self-repair techniques are used for the better yield of the system by using various techniques like 1-D Redundancy and 2-D Redundancy. Test, Redundancy analysis, Repair delivery are the three basic steps for the memory repair. A built in redundancy algorithms (BIRA) are used to implement built in self-repair (BISR).