Abstract: Graph neural networks (GNNs)
have recently exploded in popularity thanks to their broad applicability
to ubiquitous graph‐related problems such as quantum chemistry, drug
discovery, and high energy physics. However, meeting demand for novel
GNN models and fast inference simultaneously is challenging because of
the gap between the difficulty in developing efficient FPGA accelerators
and the rapid pace of the creation of new GNN models. Prior art focuses
on the acceleration of specific classes of GNNs but lacks the
generality to work across existing models or to extend to new and
emerging GNN models. In this work, we propose a generic GNN acceleration
framework using High‐Level Synthesis (HLS), named GenGNN, with two‐fold
goals. First, we aim to deliver ultra‐fast GNN inference without any
graph pre‐processing for real‐time requirements. Second, we aim to
support a diverse set of GNN models with the extensibility to flexibly
adapt to new models. The framework features an optimized message‐passing
structure applicable to all models, combined with a rich library of
model‐specific components. We verify our implementation on‐board on the
Xilinx Alveo U50 FPGA and observe a speed‐up of up to 25x against CPU
(6226R) baseline and 13x against GPU (A6000) baseline.
Bio:
Cong (Callie) Hao is an assistant professor in ECE at Georgia Tech,
where she currently holds the Sutterfield Family Early Career
Professorship. She was a postdoctoral fellow in ECE at Georgia Tech from
2020-2021 and also worked as a postdoctoral researcher in ECE at the
University of Illinois at Urbana-Champaign from 2018-2020. She received
the Ph.D. degree in Electrical Engineering from Waseda University in
2017, and the M.S. and B.S. degrees in Computer Science and Engineering
from Shanghai Jiao Tong University. Her primary research interests lie
in the joint area of efficient hardware design and machine learning
algorithms, as well as reconfigurable and high-efficiency computing and
electronic design automation.