基于GNN的拥堵预测的可通用交叉图嵌入
目前,随着技术节点的扩展,在早期设计阶段的准确预测模型可以大大缩短设计周期。特别是在逻辑综合阶段,预测由于不恰当的逻辑组合而导致的单元拥堵可以减少后续物理实现的负担。已有人尝试使用图形神经网络(GNN)技术来解决逻辑综合阶段的拥塞预测问题。然而,由于GNN的核心思想是建立在消息传递框架之上的,因此它们需要有丰富的单元特征来实现合理的性能,这在早期的逻辑综合阶段是不现实的。为了解决这个限制,我们提出了一个框架,可以直接学习给定网表的嵌入,以提高我们节点特征的质量。流行的基于随机行走的嵌入方法,如Node2vec、LINE和DeepWalk,都存在跨图对齐的问题,而且对未见过的网表图的通用性较差,产生了较差的性能并耗费了大量的运行时间。在我们的框架中,我们引入了一个卓越的替代方案,以获得节点嵌入,并使用矩阵分解方法在网表图中进行推广。我们提出了一种高效的子图层面的小批量训练方法,可以保证并行训练并满足大规模网表的内存限制。我们展示了利用开源EDA工具(如DREAMPLACE和OPENROAD框架)在各种公开的电路上的结果。通过将网表上的学习嵌入与GNN结合起来,我们的方法提高了预测性能,可以通用于新的电路线,并且训练效率高,可能会节省90%以上的运行时间。
原文题目:Generalizable Cross-Graph Embedding for GNN-based Congestion Prediction
原文:Presently with technology node scaling, an accurate prediction model at early design stages can significantly reduce the design cycle. Especially during logic synthesis, predicting cell congestion due to improper logic combination can reduce the burden of subsequent physical implementations. There have been attempts using Graph Neural Network (GNN) techniques to tackle congestion prediction during the logic synthesis stage. However, they require informative cell features to achieve reasonable performance since the core idea of GNNs is built on the message passing framework, which would be impractical at the early logic synthesis stage. To address this limitation, we propose a framework that can directly learn embeddings for the given netlist to enhance the quality of our node features. Popular random-walk based embedding methods such as Node2vec, LINE, and DeepWalk suffer from the issue of cross-graph alignment and poor generalization to unseen netlist graphs, yielding inferior performance and costing significant runtime. In our framework, we introduce a superior alternative to obtain node embeddings that can generalize across netlist graphs using matrix factorization methods. We propose an efficient mini-batch training method at the sub-graph level that can guarantee parallel training and satisfy the memory restriction for large-scale netlists. We present results utilizing open-source EDA tools such as DREAMPLACE and OPENROAD frameworks on a variety of openly available circuits. By combining the learned embedding on top of the netlist with the GNNs, our method improves prediction performance, generalizes to new circuit lines, and is efficient in training, potentially saving over 90% of runtime.
相关文章
- 在 Go 里用 CGO?这 7 个问题你要关注!
- 9款优秀的去中心化通讯软件 Matrix 的客户端
- 求职数据分析,项目经验该怎么写
- 在OKR中,我看到了数据驱动业务的未来
- 火山引擎云原生大数据在金融行业的实践
- OpenHarmony富设备移植指南(二)—从postmarketOS获取移植资源
- 《数据成熟度指数》报告:64%的企业领袖认为大多数员工“不懂数据”
- OpenHarmony 小型系统兼容性测试指南
- 肯睿中国(Cloudera):2023年企业数字战略三大趋势预测
- 适用于 Linux 的十大命令行游戏
- GNOME 截图工具的新旧截图方式
- System76 即将推出的 COSMIC 桌面正在酝酿大变化
- 2GB 内存 8GB 存储即可流畅运行,Windows 11 极致精简版系统 Tiny11 发布
- 迎接 ecode:一个即将推出的具有全新图形用户界面框架的现代、轻量级代码编辑器
- loongarch架构介绍(三)—地址翻译
- Go 语言怎么解决编译器错误“err is shadowed during return”?
- 敏捷:可能被开发人员遗忘的部分
- Denodo预测2023年数据管理和分析的未来
- 利用数据推动可持续发展
- 在 Vue3 中实现 React 原生 Hooks(useState、useEffect),深入理解 React Hooks 的