Reconfigurable ECC for adaptive protection of memory

Abstract

Post-silicon healing techniques that rely on built-in redundancy (e.g. row/column redundancy) remain effective in healing manufacturing defects and process variation induced failures in nanoscale memory. They are, however, not effective in improving robustness under various run-time failures. Increasing run-time failures in memory, specifically in case of low-voltage low-power memory, has emerged as a major design challenge. Traditionally, a uniform worst-case protection using Error Correction Code (ECC) is used for all blocks in a large memory array for runt-time error resiliency. However, with both spatial and temporal shift in intrinsic reliability of a memory block, such uniform protection can be unattractive in terms of either ECC overhead or protection level. We propose a novel Reconfigurable ECC approach, which can adapt, in space and time, to varying reliability of memory blocks by using an ECC that can provide the right amount of protection for a memory block at a given time. We show that such an approach is extremely effective in diverse applications.