Artificial intelligence (AI) is reshaping society, science, and computing—from scientific discovery and advanced manufacturing to health care, communications, and intelligent-edge systems. Yet the escalating computational and energy demands of frontier AI are exposing significant limits to its efficiency, sustainability, accessibility, and scalability. Realizing AI’s full potential will require more than increasingly capable models; it will require radically new approaches to how AI systems are designed, built, and deployed.
Jointly organized by the NSF Phase-II POSE HLS4ML Project, the Discovery Partners Institute, and the NSF Accelerated AI Algorithms for Data-Driven Discovery Institute , with participation from the FastML Foundation, this two-day workshop will bring together researchers, engineers, open-source developers, and industry partners to explore new directions for the future of AI computing.
A central theme is convergent intelligence: an emerging paradigm for advancing AI in which algorithms, computing platforms, hardware and software architectures, design tools, and data resources are developed in close coordination. The workshop will explore how this integrated approach can help enable more efficient, sustainable, accessible, and scalable AI.