DGNN-Booster：动态图神经网络推理的通用 FPGA 加速器框架（英文版）

英文标题：DGNN-Booster: A Generic FPGA Accelerator Framework For Dynamic Graph Neural Network Inference中文摘要：本文提出了一个新颖的 FPGA 加速器框架 DGNN-Booster，旨在解决动态图神经网络分析中的硬件部署问题。通过基于高层次综合的实时 DGNN 推断，DGNN-Booster 可以实现最多 5.6 倍的加速比，比 CPU、GPU 基线分别高 8.4 倍和 1000 倍以上。英文摘要：Dynamic Graph Neural Networks (DGNNs) are becoming increasingly popular dueto their effectiveness in analyzing and predicting the evolution of complexinterconnected graph-based systems. However, hardware deployment of DGNNs stillremains a challenge. F