Font Size: a A A

Self-healing cellular automata to correct soft errors in defective embedded program memories

Posted on:2010-01-17Degree:M.SType:Thesis
University:Utah State UniversityCandidate:Voddi, VarunFull Text:PDF
GTID:2448390002983915Subject:Engineering
Abstract/Summary:
Static Random Access Memory (SRAM) cells in ultra-low power Integrated Circuits (ICs) based on nanoscale Complementary Metal Oxide Semiconductor (CMOS) devices are likely to be the most vulnerable to large-scale soft errors. Conventional error correction circuits may not be able to handle the distributed nature of such errors and are susceptible to soft errors themselves. In this thesis, a distributed error correction circuit called Self-Healing Cellular Automata (SHCA) that can repair itself is presented. A possible way to deploy a SHCA in a system of SRAM-based embedded program memories (ePM) for one type of chip multi-processors is also discussed. The SHCA is compared with conventional error correction approaches and its strengths and limitations are analyzed.
Keywords/Search Tags:Soft errors, SHCA
Related items