Electronic Design Automation and Co-Design
Electronic Design Automation and Co-Design at the University of Maryland focuses on developing advanced computational tools and integrated design methodologies that enable efficient creation, simulation, and optimization of complex electronic systems. Research in this area spans circuit and system-level modeling, hardware–software co-design, design for reliability and manufacturability, and AI-driven automation for semiconductor and heterogeneous integration platforms. The goal is to streamline the design cycle, enhance performance and energy efficiency, and ensure seamless collaboration between hardware and software components across diverse applications—from embedded systems to large-scale data centers.
Co-Design for Heterogeneous Integration at the University of Maryland focuses on creating comprehensive, cross-disciplinary design frameworks that tightly couple materials, devices, circuits, and system architectures to advance the next generation of semiconductor technologies. This area bridges traditionally separate domains—such as chip design, packaging, thermal management, and system integration—to enable optimized performance, power efficiency, and reliability in increasingly complex and miniaturized systems.
Researchers develop methodologies that consider electrical, thermal, and mechanical interactions from the earliest stages of design, allowing for predictive modeling and co-optimization across multiple technology layers. The work emphasizes seamless integration of diverse technologies—such as compound semiconductors, silicon, photonics, and advanced interconnects—into unified platforms.
By leveraging electronic design automation (EDA), multi-physics simulation, and AI-driven optimization, co-design for heterogeneous integration enables innovative architectures for high-performance computing, data centers, autonomous systems, and emerging applications in AI, communications, and sensing. This holistic approach ensures that materials and device innovations can be effectively translated into scalable, system-level solutions.
AI-Based Electronic Design Automation (EDA) Tools at the University of Maryland focus on harnessing artificial intelligence and machine learning to revolutionize the design, optimization, and verification of semiconductor devices and systems. This research integrates data-driven algorithms with traditional EDA workflows to significantly accelerate design cycles, enhance accuracy, and enable intelligent decision-making throughout the semiconductor design process.
By embedding AI and machine learning into key stages—such as layout generation, placement and routing, power and thermal analysis, and design verification—researchers are developing adaptive tools that learn from prior designs and predict optimal configurations in real time. These tools automate complex design tasks, identify potential reliability or performance issues early, and enable co-optimization across electrical, thermal, and mechanical domains.
Efforts also focus on AI-enhanced multi-physics modeling, design-space exploration, and hardware–software co-design, where intelligent algorithms guide trade-offs between power, performance, and area (PPA). The use of reinforcement learning, graph neural networks, and large-scale data analytics is driving breakthroughs in chip architecture generation and system-level synthesis.
Ultimately, AI-based EDA tools aim to transform semiconductor design from a manual, iterative process into an intelligent, autonomous, and collaborative ecosystem—supporting faster innovation cycles and advancing heterogeneous integration, advanced packaging, and next-generation computing technologies.
EDA for Novel Materials, Devices, and Circuits at the University of Maryland develops next-generation electronic design automation tools and methodologies that bridge the gap between emerging materials science and advanced circuit design. As semiconductor technology extends beyond traditional silicon, this research area enables accurate modeling, simulation, and optimization of devices and systems built from novel materials such as wide-bandgap semiconductors, 2D materials, ferroelectrics, spintronic materials, and beyond-CMOS architectures.
Researchers focus on creating physics-based compact models and multi-scale simulation frameworks that capture the unique electrical, thermal, and mechanical behavior of these new materials and device structures. These models are integrated into circuit- and system-level design environments, allowing for predictive co-design across materials, devices, and architectures.
The work also explores EDA-driven design–technology co-optimization (DTCO), linking material properties and fabrication processes directly to system performance metrics such as power efficiency, reliability, and signal integrity. Advanced computational techniques, including AI and machine learning, are applied to accelerate device characterization, automate model extraction, and optimize circuit topologies that exploit the distinct advantages of novel technologies.
Through these innovations, the University of Maryland’s research in EDA for Novel Materials, Devices, and Circuits supports the realization of next-generation electronics—from ultra-efficient power devices and high-frequency communication systems to neuromorphic and quantum-inspired architectures—enabling the continued scaling and diversification of semiconductor technologies.
