Lookup table-based negative-bias temperature instability effect and leakage power co-optimization using genetic algorithm approach,International Journal of Circuit Theory and Applications

当前位置： X-MOL 学术 › Int. J. Circ. Theory Appl. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

Lookup table-based negative-bias temperature instability effect and leakage power co-optimization using genetic algorithm approach
International Journal of Circuit Theory and Applications ( IF 1.8 ) Pub Date : 2021-05-03 , DOI: 10.1002/cta.3038
Abhishek Bhattacharjee ₁ , Dheeraj Kumar Sahu ₁ , Sambhu Nath Pradhan ₁

Affiliation

As a result of continuous scaling of transistors, negative bias temperature instability (NBTI) has become a serious reliability concern which causes the device to degrade over its lifetime. Similarly, leakage power also remains a key issue in deep submicron technologies. Both NBTI-related delay degradation and standby time leakage power depend strongly on the circuit's input patterns. The input vector control (IVC) technique can be used for mitigating both NBTI-related delay degradation and leakage power. However, the input vectors with the least NBTI-induced degradation may not be the same ones with the least leakage power in any circuit; therefore, co-optimization is needed to minimize leakage and NBTI-induced degradation. In this paper, a genetic algorithm (GA)-based approach is presented to co-optimize both NBTI-related performance degradation and also leakage power. We are also presenting a trade-off analysis by giving different weightage to the NBTI effect and leakage power. Simulation results applied on ISCAS'85 benchmark circuits illustrate that on average our proposed technique saves up to 21.85% and 17.68% circuit performance degradation and the leakage power, respectively.

中文翻译：

基于查找表的负偏置温度不稳定性效应和使用遗传算法的泄漏功率协同优化

由于晶体管的连续缩放，负偏置温度不稳定性 (NBTI) 已成为严重的可靠性问题，这会导致器件在其使用寿命期间性能下降。同样，泄漏功率也仍然是深亚微米技术的关键问题。NBTI 相关的延迟退化和待机时间泄漏功率都在很大程度上取决于电路的输入模式。输入矢量控制 (IVC) 技术可用于减轻与 NBTI 相关的延迟退化和泄漏功率。然而，在任何电路中，NBTI 引起的退化最小的输入向量可能与泄漏功率最小的输入向量不同；因此，需要协同优化以最大限度地减少泄漏和 NBTI 引起的退化。在本文中，提出了一种基于遗传算法 (GA) 的方法来共同优化与 NBTI 相关的性能下降和泄漏功率。我们还通过对 NBTI 效应和泄漏功率赋予不同的权重来进行权衡分析。应用于 ISCAS'85 基准电路的仿真结果表明，我们提出的技术平均分别节省了 21.85% 和 17.68% 的电路性能下降和泄漏功率。

更新日期：2021-07-05

点击分享查看原文

点击收藏

阅读更多本刊最新论文本刊介绍/投稿指南11