Session

· Contributors · Organizations · Search Program · Flagged · Happening Now

Research Manuscript: AI for Analog Synthesis

Session Chair

DescriptionAI has become a major underlying method for solving the analog synthesis problem. This session tackles the multi-flavored challenges of analog synthesis using various modern AI techniques. They include hierarchical, multi-agent-based approaches, reinforcement learning, transfer learning, and advanced regression methods. The goals range from exploring Pareto fronts to considering PVT corners in the synthesis.

Event TypeResearch Manuscript

TimeThursday, June 2710:30am - 12:00pm PDT

Location3002, 3rd Floor

Topics

Keywords

Presentations

10:30am - 10:45am PDT	HiMOSS: A Novel High-dimensional Multi-objective Optimization Method via Adaptive Gradient-Based Subspace Sampling for Analog Circuit Sizing Authors Tianchen Gu Ruiyu Lyu Zhaori Bi Changhao Yan Fan Yang Dian Zhou Tao Cui Xin Liu Zaikun Zhang Xuan Zeng
10:45am - 11:00am PDT	EVDMARL: Efficient Value Decomposition-based Multi-Agent Reinforcement Learning with Domain-Randomization for Complex Analog Circuit Design Migration Authors Handa Sun Zhaori Bi Wenning Jiang Ye Lu Changhao Yan Fan Yang Wenchuang Hu Sheng-Guo Wang Dian Zhou Xuan Zeng
11:00am - 11:15am PDT	Using Probabilistic Model Rollouts to Boost the Sample Efficiency of Reinforcement Learning for Automated Analog Circuit Sizing Authors Mohsen Ahmadzadeh Georges Gielen
11:15am - 11:30am PDT	Learn-by-Compare: Analog Performance Prediction using Contrastive Regression with Design Knowledge Authors Zihu Wang Karthik Somayaji N.S. Peng Li
11:30am - 11:45am PDT	PVTSizing: A TuRBO-RL-Based Batch-Sampling Optimization Framework for PVT-Robust Analog Circuit Synthesis Authors Zichen Kong Xiyuan Tang Wei Shi Yiheng Du Yibo Lin Yuan Wang
11:45am - 12:00pm PDT	KATO: Knowledge Alignment And Transfer for Transistor Sizing Of Different Design and Technology Authors Wei Xing Weijian Fan Zhuohua Liu Yuan Yao Yuanqi Hu

DAC 2024