On-line characterization and reconfiguration for single event upset variations

Abstract

The amount of physical variation among electronic components on a die is increasing rapidly. There is a need for a better understanding of variations in transient fault susceptibility, and for methods of on-line adaptation to such variations. We address three key research questions in this area. First, we investigate accelerated characterization of individual latch susceptibilities. We find that on the order of 10 upsets per latch must be observed for variations to be adequately characterized. Second, we propose a method of on-line hardware reconfiguration using incremental place-and-route on FPGAs. Surprisingly, we find that highly localized place-and-route changes (e.g. restricted to groups of 8 flip-flops) are sufficient for realizing most of the possible benefits. Lastly, we quantify potential improvements in system-level soft error rates via Monte Carlo simulation experiments. The study highlights both what is required for and what can be gained by on-line adaptation.

Self-Test and Adaptation for Random Variations in Reliability

“…Some methods have been proposed for mapping marginal cells in an SRAM array and for preferentially matching an application to the appropriate cells/regions [8], but these are not applicable to logic. We previously performed an initial study of self-adaptation given random variations in latch reliability [9]. In that work, we established some limits on self-test, and determined that in many cases lower-cost methods will be needed.…”

“…Transient faults are one-time events that upset the state of a node and potentially lead to system-level soft errors [9]. Faults can be triggered by particle radiation, coupling, thermal noise [19], and other mechanisms.…”

“…In previous work, we formulated a similar problem by making a simplification: flip-flops were assigned a scalar weight, and the cost of matching elements i and j was a simple product of the two weights [9]. In fact, this type of model is a recurring one seen in analogous settings [8] [13].…”