Rate My Professor Matthew Guthaus

Guthaus's research encompasses electronic design automation enhanced by machine learning and graph neural networks, such as GATMesh for clock mesh timing analysis, GAT-Steiner for routing prediction, and effective capacitance modeling. He has developed open-source memory compilers including OpenRAM, a framework for random-access memories jointly with Oklahoma State University, and OpenGCRAM for gain cell architectures optimized for AI workloads. Additional areas include current-mode clocking to minimize power and switching noise, resonant and charge-recovery clocking techniques like distributed LC resonant grids and switched capacitor quasi-adiabatic clocks, and electromagnetic interference mitigation through dynamic EMI shifting and clock tree optimization. Key publications comprise “OpenRAM: An Open-Source Memory Compiler” (ICCAD 2016), “GAT-Steiner: Rectilinear Steiner Minimal Tree Prediction Using GNNs” (ICCAD 2024), “GATMesh: Clock Mesh Timing Analysis using Graph Neural Networks” (arXiv 2025), “Distributed LC Resonant Clock Grid Synthesis” (TCAS-I 2012), and “HCDN: Hybrid-Mode Clock Distribution Networks” (TCAS-I 2018). He received a Google Faculty Research Award in 2020 for computer memory and chip design promoting open-source hardware, led an NSF steering committee report in 2022 on hardware workforce revitalization, and contributed to UCSC's record patent awards, including two personal patents in 2019. Guthaus delivered a talk on OpenROAD’s evolving optimizer at ORConf 2025.