Jinghan Yao

Jinghan Yao works on high-performance computing systems for large-scale machine learning, focusing on how to make distributed training and inference of large models (like transformers and language models) faster and more efficient across clusters of GPUs. Much of the recent work involves designing communication libraries and network-aware strategies (such as MPI collective operations and GPU interconnect balancing) that help many computers work together efficiently when training huge AI models. The group also has earlier work on computer vision tasks like image segmentation and efficient attention mechanisms in neural networks.

Publication output was minimal before 2019, rose modestly through 2021, then increased substantially from 2023 onward, with a notably high count projected for 2026, suggesting a growing and increasingly active research pace in recent years.

Generated by claude-sonnet-5 from public bibliographic data · Jul 20, 2026