Redundancy algorithm using line based method for memories achieving less area overhead

Abstract

Test cost and yield improvement is becoming important parameters with the growth of memory capacity and density. Built-in redundancy analysis (BIRA) is popularly used for embedded memories to solve yield and quality issues by removing faulty cells with available goods cells. Different BIRA approaches require different area overheads to get optimal repairs. It is difficult to get low area overhead and at the same time optimal repair rate. A new BIRA method is proposed and it uses a line-based search operation. The new BIRA reduces the storage capacity for storing faulty address data by ignoring the data of unnecessary faulty cell addresses. By using a line fail count comparison method, the proposed BIRA analyzes redundant spares quickly to get optimal repair rate. By simulation results it is verified that the proposed BIRA does the spare allocations with low area overhead.