Research

We work across computer architecture, with a special focus on the performance, reliability, and security of GPUs, heterogeneous systems, and accelerators. Our approach pairs careful workload characterization with new architectural mechanisms and fast, accurate modeling tools. Our work spans:

GPU and Multi-GPU Architecture

Memory hierarchies, caches, interconnects, and bandwidth management for GPUs and multi-GPU systems.

Performance Modeling and Simulation

Fast, accurate prediction and simulation of AI workloads on GPUs, multi-GPU systems, and emerging hardware.

Accelerating Emerging Workloads

Architectural and system support for automata processing and other emerging applications on GPUs and accelerators.

See all publications.

Our research has been supported by the National Science Foundation (including an NSF CAREER award), the Department of Energy, Google, and NVIDIA. Our students have interned at AMD, Intel, and Pacific Northwest National Laboratory, and our Ph.D. alumni have gone on to faculty positions and to industry roles at AMD and NVIDIA.

People

Current Members

Ph.D. Alumni

Current Collaborators

Master's and Undergraduate Alumni

Undergraduate Researchers

  • B. RuthUVA NSF REU Project 2025
  • W. KaiserUVA NSF REU Project 2024–2025
  • P. TangellaUVA DOE Project 2024–2025
  • S. KapaUVA DOE Project 2024
  • H. NguyenUVA Dean's Fellowship 2024
  • A. ChungW&M Charles Center Summer Research Scholar 2020
  • S. HongW&M Charles Center Summer Research Scholar 2018
  • C. WeinshenkerW&M Honors Thesis · Park Award winner 2017
  • R. RisqueW&M Senior Thesis 2016

Master's Students

  • B. DaceyUVA2025
  • R. XingW&M2019
  • C. AmesW&M2017
  • R. ChaiW&M2017
  • F. LuoW&M2016
  • H. WangW&M2016

Recent News

All news →