ISOP: Machine Learning-Assisted Inverse Stack-Up Optimization for Advanced Package Design
ISOP: Machine Learning-Assisted Inverse Stack-Up Optimization for Advanced Package Design
复制标题
DOI：
10.23919/date56975.2023.10137055
复制
发表时间：
2023-04
期刊：
影响因子：
--
通讯作者：
Hyunsu Chae;B. Mutnury;Keren Zhu;D. Wallace;D. Winterberg;D. D. Araujo-D.;J. Reddy;Adam R. Klivans;D. Pan
中科院分区：
文献类型：
--
作者：
Hyunsu Chae;B. Mutnury;Keren Zhu;D. Wallace;D. Winterberg;D. D. Araujo-D.;J. Reddy;Adam R. Klivans;D. Pan
未来的计算需要异构集成，例如，最近采用了小芯片方法。然而，高速跨芯片互连和封装对于整个系统性能至关重要。作为先进封装的示例，高密度互连（HDI）印刷电路板（PCB）已广泛用于从手机到计算服务器的复杂电子产品中。现代HDI PCB可以具有超过20层，每层具有其独特的材料特性和几何尺寸，即，堆叠，以满足各种设计约束和性能优化。然而，层叠设计通常在工业中手动完成，其中有经验的设计师可能花费许多时间来调整物理尺寸和材料以满足所需的规格。然而，这个过程是耗时的，乏味的，并且是次优的，这在很大程度上取决于设计师的专业知识。在本文中，我们提出了一个新的框架，ISOP，使用机器学习的逆堆叠优化先进的包装设计自动化的堆叠设计。给定目标设计规格，ISOP会自动搜索理想的层叠设计参数，同时优化性能。我们开发了一种新的机器学习辅助超参数优化方法，使搜索高效可靠。实验结果表明，ISOP是更好的品质因数（FoM）比传统的模拟退火和贝叶斯优化算法，与我们所有的设计目标，满足更短的运行时间。我们还将我们的全自动ISOP与行业内的专家设计师进行了比较，并取得了非常有希望的结果，周转时间减少了几个数量级。
Future computing calls for heterogeneous integration, e.g., the recent adoption of the chiplet methodology. However, high-speed cross-chip interconnects and packaging shall be critical for the overall system performance. As an example of advanced packaging, a high-density interconnect (HDI) printed circuit board (PCB) has been widely used in complex electronics from cell phones to computing servers. A modern HDI PCB may have over 20 layers, each with its unique material properties and geometrical dimensions, i.e., stack-up, to meet various design constraints and performance optimizations. However, stack-up design is usually done manually in the industry, where experienced designers may devote many hours to adjusting the physical dimensions and materials to meet the desired specifications. This process, however, is time-consuming, tedious, and sub-optimal, largely depending on the designer's expertise. In this paper, we propose to automate the stack-up design with a new framework, ISOP, using machine learning for inverse stack-up optimization for advanced package design. Given a target design specification, ISOP automatically searches for ideal stack-up design parameters while optimizing performance. We develop a novel machine learning-assisted hyper-parameter optimization method to make the search efficient and reliable. Experimental results demonstrate that ISOP is better in figure-of-merit (FoM) than conventional simulated annealing and Bayesian optimization algorithms, with all our design targets met with a shorter runtime. We also compare our fully-automated ISOP with expert designers in the industry and achieve very promising results, with orders of magnitude reduction of turn-around time.