An FPGA-Based Solution for a Graph Neural Network Accelerator

GNN have very high requirements for computing power and memory, and the software implementation of GNN does not meet performance targets. As a result, there is an urgent need for hardware-based GNN acceleration. While traditional convolutional neural network (CNN) hardware acceleration has many solutions, the hardware acceleration of GNN has not been fully discussed and researched. This white paper will review the latest GNN algorithms, current status of acceleration technology research and a discussion of FPGA-based GNN acceleration technology.