Analyzing Scaled Reduced Precision Redundancy for Error Mitigation Under Proton Irradiation
- 28 January 2022
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Nuclear Science
- Vol. 69 (7) , 1485-1491
Abstract
Reduced precision redundancy (RPR) is an alternative to triple modular redundancy (TMR) that reduces the area overhead at the expense of minor accuracy loss in case of error. In this work, we propose a Scaled RPR approach for multistage circuits and analyze the error mitigation tradeoffs. As a case study, several fast Fourier transform designs were tested with low-energy protons and fault injection. Experimental results show that the proposed approach achieves error mitigation with good accuracy, while significantly reducing the area overhead with respect to a full precision TMR approach.Keywords
Funding Information
- Spanish Ministry of Science and Innovation (PID2019-106455GB-C21)
- Community of Madrid (49.520608.9.18)
This publication has 15 references indexed in Scilit:
- Strategies for Removing Common Mode Failures From TMR Designs Deployed on SRAM FPGAsIEEE Transactions on Nuclear Science, 2018
- Partial TMR in FPGAs Using Approximate Logic CircuitsIEEE Transactions on Nuclear Science, 2016
- Reduced-Precision Redundancy on FPGAsInternational Journal of Reconfigurable Computing, 2011
- Fine-Grain SEU Mitigation for FPGAs Using Partial TMRIEEE Transactions on Nuclear Science, 2008
- A Comparison of TMR With Alternative Fault-Tolerant Design Techniques for FPGAsIEEE Transactions on Nuclear Science, 2007
- Energy-efficient soft error-tolerant digital signal processingIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2006
- Designing fault-tolerant techniques for SRAM-based FPGAsIEEE Design & Test of Computers, 2004
- Soft error rate mitigation techniques for modern microcircuits2003
- Reduced precision redundancy for low-power digital filtering2001
- Pipeline and Parallel-Pipeline FFT Processors for VLSI ImplementationsIEEE Transactions on Computers, 1984