Suggested Single Event Upset Figure of Merit

Prevention of Single Event Upsets in Microelectronics

“…This is, obviously, a measure of the relative change in threshold for single-event upset for future technologies; it is not analogous to the "figure of merit" [90] which predicts upsets per bit-day. This sensitivity measure, defined as 11S, depends on the scaling of Qi, to Qo,.…”

“…Again, this is a quantitative representation, but it shows that non-scaling parameters all conspire to increase the sensitivity of future technologies to single-event upset. Figure 39 shows the effect of constant field scaling on the figure of merit for single-event error rate [90]: when constant field scaling rules are applicable, error rate remains constant. This conclusion was also reached in previous studies on scaling [88,89].…”

“…As the alternative form for the figure of merit [90] uses the equivalency Qc ,LET (where the variable c is the smallest of the sensitive volume dimensions, intended to be the thickness of the sensitive junction). This equivalency indicates that Q.…”

“…An approximate expression developed by Petersen, et al, [123] which describes the SEU error rate in errors/bit-day is…”

“…The first works on the scaling of single-event effects [88,89] were based on the acceptance of constant-field scaling for future devices and on a "figure of merit" [90] for assessing the vulnerability of ICs, given information on their feature size and total sensitive area; funneling was "patched" onto these models as an estimated extension in sensitive volume.…”