Paper Info
Title
FP-DNN: An Automated Framework for Mapping Deep Neural Networks onto FPGAs with RTL-HLS Hybrid Templates
Abstract
DNNs (Deep Neural Networks) have demonstrated great success in numerous applications such as image classification, speech recognition, video analysis, etc. However, DNNs are much more computation-intensive and memory-intensive than previous shallow models. Thus, it is challenging to deploy DNNs in both large-scale data centers and real-time embedded systems. Considering performance, flexibility, and energy efficiency, FPGA-based accelerator for DNNs is a promising solution. Unfortunately, conventional accelerator design flows make it difficult for FPGA developers to keep up with the fast pace of innovations in DNNs. To overcome this problem, we propose FP-DNN (Field Programmable DNN), an end-to-end framework that takes TensorFlow-described DNNs as input, and automatically generates the hardware implementations on FPGA boards with RTL-HLS hybrid templates. FP-DNN performs model inference of DNNs with our high-performance computation engine and carefully-designed communication optimization strategies. We implement CNNs, LSTM-RNNs, and Residual Nets with FPDNN, and experimental results show the great performance and flexibility provided by our proposed FP-DNN framework.
Year
DOI
Venue
2017
10.1109/FCCM.2017.25
2017 IEEE 25th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM)
Keywords
Field
DocType
Deep Neural Networks,FPGA,Automation,RTL-HLS
Kernel (linear algebra),Data modeling,Efficient energy use,Computer science,Parallel computing,High-level synthesis,Field-programmable gate array,Real-time computing,Automation,Design flow,Artificial neural network
Conference
ISBN
Citations 
PageRank 
978-1-5386-4038-8
42
1.79
References 
Authors
16
9
Name
Order
Citations
PageRank
Yijin Guan143818.66
Hao Liang2545.83
Ning-Yi Xu356336.18
Wen-Qiang Wang4754.89
Shaoshuai Shi51716.85
Xi Chen622649.58
Guangyu Sun71920111.55
Wei Zhang875166.85
Jason Cong97069515.06