Project Details
Projekt Print View

MOTARO: A modular approach for test and repair of embedded non-volatile memories.

Subject Area Computer Architecture, Embedded and Massively Parallel Systems
Electronic Semiconductors, Components and Circuits, Integrated Systems, Sensor Technology, Theoretical Electrical Engineering
Term from 2013 to 2018
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 242177335
 
Final Report Year 2017

Final Report Abstract

The main driving force of the research conducted was addressing reliability challenges of existing and emerging NVMs implemented in embedded systems. This was achieved by developing novel memory repair techniques that enable wider adoption of NVMs in digital systems. Each investigated technique focuses on different reliability aspects of NVMs, i.e.: Word-level repair provides on-line repair of wear-out memory cells that can occur during memory lifetime. - Block-level repair focuses on post-production faults in the memory array. - Error-correcting code with increased hard error correction capability provides protection against hard faults that can occur in the memory array and soft errors caused e.g., by external factors. - Software-based program adaptation techniques allow to reuse faulty program memories for storing program code. - Configuration data protection mechanism based on N-modular redundancy allows to reliably store important data even in the corrupted memory areas. All proposed techniques were evaluated, and their repair capabilities measured. Moreover, it was shown that by combining the proposed techniques into a consistent system significantly improved memory repair can be achieved. The applicability of the investigated mechanisms for embedded systems was proven through system emulations. In addition, hardware-based techniques were implemented in ASIC resulting in 186 produced chips. In order to focus entirely on the memory repair aspect we abandoned aspects related to memory built-in self-test mechanisms.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung