GridNetOpt: Fast Full-Chip EM-Aware Power Grid Optimization Accelerated by Deep Neural Networks
GridNetOpt: Fast Full-Chip EM-Aware Power Grid Optimization Accelerated by Deep Neural Networks
复制标题
DOI：
10.1109/tcad.2022.3206397
复制
发表时间：
2023-05
影响因子：
2.9
通讯作者：
Han Zhou;Yibo Liu;Wentian Jin;S. Tan
中科院分区：
文献类型：
--
作者：
Han Zhou;Yibo Liu;Wentian Jin;S. Tan
本文提出了一种名为GridNetOpt的快速全芯片电迁移（EM）感知且受IR压降约束的优化框架，用于片上电网网络，该框架由深度神经网络（DNN）加速。与现有的基于线性规划的方法相比，新方法采用了更灵活的基于共轭梯度的优化来确定电网导线宽度。为了降低在每次迭代步骤中使用全芯片IR压降分析计算伴随网络灵敏度的高昂成本，灵敏度通过训练好的条件生成对抗网络（CGAN）来计算。新方法利用DNN的可微特性实现快速灵敏度计算。灵敏度，即节点电压相对于导线电阻的变化，将在优化过程中指导搜索方向。为了更准确地考虑EM失效影响，训练数据取自不同导线宽度和电流负载下的电网，这些电网由最先进的基于全芯片多物理场的EM - IR压降耦合分析工具进行分析。这与现有的基于线性规划的方法形成对比，后者只能处理无失效导线或具有非零电阻的导线。对来自ARM Cortex - M0处理器设计的多个合成电网基准测试的数值结果表明，所提出的GridNetOpt相较于使用传统伴随网络方法的基于共轭梯度的方法，至少能实现一个数量级的加速。与之前使用GridNet的局部电网修复工作相比，GridNetOpt在我们测试的所有基准测试中导致的面积开销更小。它还能降低含无失效导线的电网电路的IR压降，而这是局部GridNet方法无法做到的。
This article presents a fast full-chip electromigration (EM) aware IR drop constrained optimization framework, named GridNetOpt, for on-chip power grid networks accelerated by deep neural networks (DNNs). Compared to the existing linear programming-based methods, the new method employs more flexible conjugate gradient-based optimization to size the wire width of the power grids. To mitigate the high cost of sensitivity calculation of the adjoint network using full-chip IR drop analysis at every iteration step, the sensitivity is computed via a trained conditional generative adversarial network (CGAN). The new method exploits the differentiable characteristics of DNNs for fast sensitivity computation. The sensitivity, which is the node voltage with respect to wire resistance, will guide the search direction during the optimization process. In order to consider more accurate EM failure effects, the training data is obtained from the power grids under different wire widths and current loads analyzed by a state-of-the-art full-chip multiphysics-based coupled EM-IR drop analysis tool. This is in contrast with the existing linear programming-based methods, in which only immortal wires or wires with nonzero resistance can be dealt with. Numerical results on a number of synthesized power grid benchmarks from ARM Cortex-M0 processor designs show that the proposed GridNetOpt can lead to at least an order of magnitude speedup over the conjugate gradient-based method using the traditional adjoint network method. Compared to the previous localized power grid fixing work with GridNet, GridNetOpt leads to smaller area overhead for all the benchmarks we tested. It can also reduce IR drops for power grid circuits with immortal wires, which is not possible with the localized GridNet method.