摘要
一位新闻记者-机器人与机器学习的工作人员新闻编辑每日新闻-机器学习的研究结果在一份新的报告中讨论。根据NewsRx记者从中国杭州发回的新闻报道，研究表明：“晶体管尺寸的不断缩小导致了电压噪声裕度的降低，并加剧了电源完整性的挑战。这一趋势加剧了人们对传统静态时序分析(STA)有效性的担忧，传统静态时序分析方法假设电源水平恒定，往往导致不精确和过于保守的结果。”本研究经费来源于国家自然科学基金(NSFC)。摘要:针对这一问题,本文提出了一种基于实时(JIT)机器学习(ML)集成的动态噪声感知STA引擎,该引擎采用Weibull累积分布函数(CDF)精确地表示动态电源噪声(PSN),并对其进行了门级表征。在输入转换时间、输出电容和三个PSN感知参数变化的情况下，评估每个定时电弧的de lay和转换时间。然后，用多层感知器(MLP)对每个定时电弧的定时进行预测，并利用特征数据进行训练。最后，结合JIT编译技术，将训练好的MLP模型集成到STA引擎中，实现了计算效率和灵活性。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Research findings on Machine Learning are discussed in a new report. According to news reporting originating from Hang zhou, People's Republic of China, by NewsRx correspondents, research stated, "Th e incessant decrease in transistor size has led to reduced voltage noise margins and exacerbated power integrity challenges. This trend intensifies concerns abo ut the efficacy of conventional static timing analysis (STA), which traditionall y assumes a constant power supply level, often resulting in imprecise and overly conservative outcomes." Financial support for this research came from National Natural Science Foundatio n of China (NSFC). Our news editors obtained a quote from the research from Zhejiang University, "T o address this, this article proposes a dynamic-noise-aware STA engine enhanced by just-in-time (JIT) machine learning (ML) integration. This approach employs t he Weibull cumulative distribution function (CDF) to accurately represent dynami c power supply noise (PSN). We perform gate-level characterization, assessing de lay and transition time for each timing arc under variations in input transition time, output capacitance, and three PSN-aware parameters. The timing for each t iming arc can then be predicted by a multilayer perceptron (MLP), trained with t he characterization data. Finally, by incorporating JIT compilation techniques, we integrate trained MLP models into the STA engine, achieving both computationa l efficiency and flexibility."
NETL
NSTL