发表机构
BNRist; Tsinghua Univ.（北京国家研究中心; 清华大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究提出Flash-CNNCap，将全矩阵电容预测转化为图像到图像回归，用密集贡献图替代标量目标，减少重建次数。经消融研究选出U-Net，在总电容和耦合精度上表现良好，速度大幅提升，还给出了部署管道的性能。
AI 中文摘要
我们提出了Flash-CNNCap，一种基于卷积神经网络（CNN）的电容提取器，它将全矩阵电容预测重新表述为空间贡献图上的图像到图像回归。先前基于标量CNN的提取器需要$O(n^2)$次前向传播来恢复具有$n$个导体的窗口中的所有成对电容。Flash-CNNCap用密集贡献图代替标量目标：一个总电容模型和一个主条件耦合模型各自预测一个空间图，通过掩码聚合将其简化为导体级值，将全矩阵重建减少到$O(n)$次传播。由此产生的总和和对称化的成对耦合在标准非对角符号约定下定义了相应的麦克斯韦式电容矩阵。这些图是从导体级标签中学习的，无需逐像素监督。对13种模型配置的消融研究选择了一个U-Net，它在总电容上与ResNet基线匹配（平均绝对相对误差为1.5-3.1%），并在所有评估的CapBench子集中实现了最强的耦合精度（平均绝对相对误差为3.0-4.6%），在平均包含134个导体的窗口上实现了$17.5\times$的全矩阵加速。一个部署的管道读取设计交换格式（DEF）几何形状并写入标准寄生交换格式（SPEF）输出，在相同基准上比OpenRCX快$4.4\times$，在51.23秒内处理1024个窗口。代码和训练模型可在此https URL上获得。
英文摘要
We present Flash-CNNCap, a CNN-based capacitance extractor that reformulates full-matrix capacitance prediction as image-to-image regression over spatial contribution maps. Prior scalar CNN-based extractors require $O(n^2)$ forward passes to recover all pairwise capacitances in a window with $n$ conductors. Flash-CNNCap replaces the scalar target with dense contribution maps: a total-capacitance model and a master-conditioned coupling model each predict a spatial map that is reduced to conductor-level values through mask aggregation, cutting full-matrix reconstruction to $O(n)$ passes. The resulting totals and symmetrized pairwise couplings define the corresponding Maxwell-style capacitance matrix under the standard off-diagonal sign convention. The maps are learned from conductor-level labels without per-pixel supervision. An ablation study over 13 model configurations selects a U-Net that matches ResNet baselines on total capacitance (1.5-3.1% MARE) and achieves the strongest coupling accuracy (3.0-4.6% MARE) across all evaluated CapBench subsets, with a $17.5\times$ full-matrix speedup on windows containing 134 conductors on average. A deployed pipeline reads Design Exchange Format (DEF) geometry and writes Standard Parasitic Exchange Format (SPEF) output, processing 1,024 windows in 51.23 seconds with a $4.4\times$ speedup over OpenRCX on the same benchmark. Code and trained models are available at https://github.com/THU-numbda/flash-cnncap.
CommentsAccepted to ICCAD-26
