FPGA-based acceleration of neural network training
Ruoyu Sang, Qiang Liu, Qi‐Jun Zhang · 2016
Neural networks (NNs) have been widely used in microwave device modeling. One of the greatest challenges is how to speed up the model training process and reduce the development cost. To address the issue, this paper exploits FPGAs to accelerate NN training. Experimental results demonstrate that the model training time can be reduced by up to 99.1%, compared to the traditional software implementation.