Search

To ensure the correct functionality of a chip throughout its entire lifetime, preliminary circuit analysis with respect to aging-induced degradation is indispensable. However, state-of-the-art techniques only allow for the consideration of uniformly applied degradations, despite the fact that different workloads will lead to different degradations due to their distinct induced activities. This imposes over-pessimism when estimating the required timing guardbands, resulting in an unnecessary loss of performance and efficiency. In this work, we propose an approach that takes real-world workload dependencies into account and generates workload-specific aging-aware standard cell libraries, allowing for accurate analysis of aging-induced degradations. We employ machine learning techniques to overcome infeasible simulation times for individual transistor aging while sustaining high prediction accuracy. We also demonstrate scalability to previously unknown workloads and discuss multiple approaches to estimate the machine learning accuracy by employing coverage metrics. In our evaluation, we achieve predictions of workload-dependent aging-aware standard cells with an average accuracy (R2 score) of 95.28%. Using predicted cell libraries in static timing analysis, timing guardbands for multiple circuits are reported with an error of less than 0.1% on average. We demonstrate that timing guardband requirements can be reduced by up to 30% when considering specific workloads over worst-case estimations as performed in state-of-the-art tool flows. Even for unknown workloads of different circuits, accurate prediction with relative errors below 1% can be achieved. [ABSTRACT FROM AUTHOR]

Aging-induced degradation imposes a major challenge to the designer when estimating timing guardbands. This problem increases as traditional worst-case corners bring over-pessimism to designers, exacerbating competitive and close-to-the-edge designs. In this work, we present an accurate machine learning approach for aging-aware cell library characterization, enabling the designer to evaluate their circuit under the impact of precisely selected degradation. Unlike state of the art, we bring cell library characterization to the designer, empowering their capability in exploring the impact of aging while protecting confidential information from the foundry at the same time. Furthermore, the fast inference of cell libraries makes it feasible, for the first time, to examine aging-induced variability analysis in a Monte-Carlo fashion. Finally, we show that the designer is able to select a less pessimistic timing guardband by choosing adequate delta threshold voltage (ΔVth) for their design and their needs. Our machine learning approach reaches an R2 score of > 99% for almost all data stored in the cell library. Only timing constraints show slightly less accuracy with an R2 score around 95%. When using ML-characterized libraries in static timing analysis, we achieve errors smaller than ± 0.5% and ± 0.1% for path delay and dynamic power, respectively. Errors in leakage power are negligible and even smaller by orders of magnitude. Our machine learning implementation for standard cell library characterization is publicly available. Download: https://opensource.mlcad.org [ABSTRACT FROM AUTHOR]