IRIS-HEP Fellow: Emily Tsai



Fellowship dates: Jan - May 2021

Home Institution: University of Texas at Austin


Project: OpenCL based implementation of graph neural networks on FPGA

Reconstructing particle trajectories, or tracking, in the CMS detector is a crucial but slow step in understanding particle collisions at the LHC. Faster tracking methods are required to keep up with the significantly increased collision rate in the future High-Luminosity LHC. This project focuses on creating an OpenCL based implementation of a graph neural network (GNN) completely on an FPGA, rather than the current implementation of the GNN on CPU and FPGA coprocessors, with the goal of speeding up tracking.

More information: My project proposal

Mentors:
  • Isobel Ojalvo (Princeton University)

  • Savannah Thais (Princeton University)

Presentations

Contact me: