Title :
Estimating the frequency threshold for logic soft errors
Author :
Mahatme, N.N. ; Gaspard, N.J. ; Jagannathan, Sarangapani ; Loveless, T.D. ; Abdel-Aziz, H. ; Bhuva, B.L. ; Massengill, Lloyd W. ; Wen, Shuli ; Wong, Rita
Author_Institution :
Vanderbilt Univ., Nashville, TN, USA
Abstract :
Logic soft errors are expected to contribute significantly towards overall soft-error failure-in-time (FIT) rates in the GHz frequency range. In this paper, the contribution of logic soft-errors, to the total soft error rate of circuits at the 28-nm technology node at high frequencies are quantified. Results suggest that the alpha particle logic soft error contribution to chip-level soft error rate is almost 50% at 0.5 GHz. The frequency threshold at which logic soft errors could exceed flip-flop errors is identified for ISCAS benchmark circuits. For most of the circuits analyzed, the frequency threshold is in the 0.9-2 GHz range. This suggests that logic soft errors could be major problem along with flip-flop errors as circuits operate at higher frequencies.
Keywords :
UHF circuits; flip-flops; frequency estimation; logic circuits; radiation hardening (electronics); FIT; ISCAS benchmark circuit; alpha particle logic soft error contribution; chip-level soft error rate; flip-flop error; frequency 0.5 GHz to 2 GHz; frequency threshold estimation; size 28 nm; soft-error failure-in-time rate; Benchmark testing; Error analysis; Flip-flops; Inverters; Logic gates; Mathematical model; Transistors; Combinational logic; Soft errors; frequency; technology scaling;
Conference_Titel :
Reliability Physics Symposium (IRPS), 2013 IEEE International
Conference_Location :
Anaheim, CA
Print_ISBN :
978-1-4799-0112-8
Electronic_ISBN :
1541-7026
DOI :
10.1109/IRPS.2013.6531991