Modeling and Experimental Demonstration of Accelerated Self-Healing Techniques

Abstract

In this paper we postulate that future electronics systems will use sleep time as an active recovery period essential for their overall performance. Our hypothesis is that by explicitly controlling the ratio of sleep vs. active and sleep conditions (e.g. higher temperatures, negative voltages), we can deeply rejuvenate electronic systems periodically to improve their metrics. We perform a series of stress and recovery experiments using commercial FPGAs to demonstrate several cases where we bring stressed chips back to within 90% of their original margin by actively rejuvenating for only 1/4 of the stress time. We validate our experiments against extracted models and present potential applications to multi-core systems.

Online Test Strategies and Optimizations for Reliable Reconfigurable Architectures

“…The application performance loss introduced by PRET depends on the test frequency and number of reconfigurable regions. In this experiment, architectures (5)(6)(7)(8)(9)(10)(11)(12)(13)(14). The performance overhead due to PORT is very low (between 0.51% and 3.73%) and scales well with higher PORT frequencies.…”

“…An FPGA-based reconfigurable fabric, manufactured in latest technology nodes (e.g., 16 nm for Xilinx' UltraScale+ family), may suffer from degradation due to aging [10,18]. Due to the increasing susceptibility of ever-shrinking nano-CMOS devices, these effects cannot be ignored anymore [11][12][13].…”

“…Profit jk = Profit intra jk + Profit inter jk (10) To calculate Profit intra jk , the average CLB stress in region k is determined as AvgStress k and then used to calculate the absolute deviation of the stress of CLB xy in region k from AvgStress k . The sum over all CLBs in region k denotes the intra-region stress imbalance.…”

“…As a result, this additive stress accumulation causes a monotonic increase in the transistor degradation over long terms. Although BTI degradation may experience a recovery effect, the recovery requires complex conditions or long relaxation periods [10] and will thus hardly affect the additive property. The monotonic and additive properties allow to consider stress during runtime (e.g., for resource management) with limited computational resources.…”

Design methodologies, frameworks and implementations on reconfigurable devices for timing and energy optimizations

“…Owing to their thermal and power sensitivity, ring oscillators have also been proposed as an appealing solution for the detection of HW Trojan threats [71,64,129,178]. Additionally, there is a number of works demonstrating the usage of ring oscillators for the evaluation and analysis of aging effects, such as Bias Temperature Instability (BTI) and Hot Carrier Injection (HCI) [143,7,113,47,174]. In this area, an interesting application where ring oscillators find place, is the detection of recycled FPGA devices via the evaluation of performance degradation caused by aging [36,132,2].…”