Unleashing Next-Generation GPU Computing for the AI Era

ABSTRACT

Modern computing platforms increasingly rely on GPUs to provide the performance, bandwidth, and memory capacity required by data-intensive workloads, from cloud and edge services to large language models (LLMs) and emerging quantum applications. In this talk, I will present our recent efforts to unlock the full potential of GPU computing. I will highlight several key innovations, including lightweight page translation invalidation, short-circuited page table walks, and flexible reconfiguration strategies for multi-instance GPUs (MIG), which together significantly improve performance and efficiency. I will also discuss our work on advancing GPU computing for edge AI. These innovations point toward the next generation of GPU systems designed to power increasingly intelligent and data-driven applications.

BIOGRAPHY

Dr. Xulong Tang is an Associate Professor in the Department of Computer Science at the University of Pittsburgh. His research spans computer architecture and systems, with a focus on GPU architectures, scalable memory management, and heterogeneous computing across high-performance, edge, and quantum systems. His group has pioneered advances in GPU memory management, temporal graph neural network training, and quantum compilation. Dr. Tang has published extensively in premier venues such as MICRO, HPCA, ISCA, ASPLOS, PLDI, ICLR, and NeurIPS. He is also active in the community, serving as General Co-Chair of ASPLOS 2026 and Vice Chair of the IEEE Micro Top Picks 2025 Selection Committee.