Planning: DCL EPSCOR: CISE Large: Hyperscale Analog Edge Computing with Brain-inspired Hardware and Learning Algorithms

University Of Oklahoma Norman Campus, Norman OK

Investigators

Abstract

Artificial intelligence (AI) technologies are transforming nearly every sector of society, yet the hardware that powers them is rapidly approaching fundamental limits in energy efficiency and scalability. Current systems, built on traditional digital CMOS architectures, suffer from the so-called von Neumann bottleneck—a separation of memory and processing that leads to significant energy and performance inefficiencies. As AI systems become more complex and pervasive, overcoming these limitations is essential to ensuring sustainable and accessible technological growth. This planning grant will support conceptualization, planning, and collaboration activities that aim to formulate new and sound plans for a large-scale project in emerging research areas. It represents an important step toward launching a transformative initiative focused on developing a new class of computing hardware that mimics the brain’s efficiency, paving the way for advances in AI performance, semiconductor technology, and workforce development. This planning grant will support the development of a comprehensive research roadmap for a novel hybrid CMOS+X in-memory analog computing framework. The project will catalyze multidisciplinary collaboration among experts in materials science, circuit design, neuroscience, and AI to co-design hardware accelerators and bio-inspired learning algorithms from the ground up. The team will organize dynamic “un-workshops,” engage with industry leaders, conduct feasibility studies on integrating emerging devices like memristors with scalable CMOS technology, and explore brain-inspired principles such as predictive coding. These activities will identify critical research questions, assess infrastructure and fabrication compatibility, and inform a future competitive CISE Large proposal. In doing so, the planning effort aims to lay the foundation for next-generation computing platforms that are energy-efficient, scalable, and aligned with national priorities in AI and semiconductor innovation. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.