Abstract

This seminar provides an overview of how modern AI systems are developed and deployed using AMD’s rapidly evolving AI ecosystem. The session introduces ROCm™, AMD’s open-source AI software platform, and explores how it enables efficient training and inference for today’s AI workloads. Participants will gain insights into recent developments in artificial intelligence and learn how AMD’s open AI ecosystem supports

applications ranging from large language models to emerging physical AI systems.

Building on this foundation, the seminar focuses on the rapid progress of Large Language Models (LLMs) and the growing impact of generative AI. Attendees will learn how ROCm integrates with industry-standard frameworks such as PyTorch to support model training, fine-tuning, and deployment. The session will also explore how advances in generative AI are transforming the way intelligent applications are developed and deployed across a wide range of domains.

The seminar further explores the emerging field of Physical AI, where AI models interact with and control robots and autonomous systems in the physical world. Participants will learn how simulation environments, synthetic data generation, and sim-to-real workflows enable robots to acquire skills efficiently before deployment in real-world environments. The discussion highlights end-to-end AI workflows that connect foundation models, robotics simulation, reinforcement learning, and real-world execution.

By bridging ROCm, large language models, and robotics applications, this seminar provides students and researchers with a comprehensive view of next-generation AI systems. Participants will gain a practical understanding of modern AI development workflows and discover how advances in generative AI and Physical AI are shaping the future of intelligent machines.

Bio.

Sonya Yang is an AI Software Engineer at AMD Research and Advanced Development (RAD), based in Taipei. She oversees academic and research collaborations across the Asia-Pacific region, including Japan, Korea, Singapore, Taiwan, Australia, and China. Sonya holds a Master’s degree in Computer Science from National Tsing Hua University and specializes in computer vision, AI models, and sim-to-real research. Her work focuses on building simulation training workflows that enable robots to perform tasks more efficiently and safely, leveraging advanced algorithms to bridge the gap between virtual and real-world performance.

All faculty and students are welcome to join.