A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
A Generic Approach for Neural Networks on FPGA
Abstract Neural networks are traditionally deployed on processor based systems. These systems suffer two major problems in terms of power consumption and time lag due to serial operation of processors. One of the major drawbacks introduced by these issues is the remote deployment of AI system. This paper addresses both issues and suggest a novel way to generically implement any neural network algorithm on a FPGA. The proposed method is tested and evaluated on Xilinx® Kintex®-7 for Radial basis function neural network with one input layer, one hidden layer and one output layer. Analysis shows a radical improvement in power consumption and improvement in the time taken to perform the same task. Actual implementation is performed in Xilinx® Vivado® software using Vivado HLS® widely used by professionals for quick implementation in the industry.
A Generic Approach for Neural Networks on FPGA
Abstract Neural networks are traditionally deployed on processor based systems. These systems suffer two major problems in terms of power consumption and time lag due to serial operation of processors. One of the major drawbacks introduced by these issues is the remote deployment of AI system. This paper addresses both issues and suggest a novel way to generically implement any neural network algorithm on a FPGA. The proposed method is tested and evaluated on Xilinx® Kintex®-7 for Radial basis function neural network with one input layer, one hidden layer and one output layer. Analysis shows a radical improvement in power consumption and improvement in the time taken to perform the same task. Actual implementation is performed in Xilinx® Vivado® software using Vivado HLS® widely used by professionals for quick implementation in the industry.
A Generic Approach for Neural Networks on FPGA
Marathe, Sameer (author)
2019-06-28
10 pages
Article/Chapter (Book)
Electronic Resource
English
A Generic Approach for Neural Networks on FPGA
| TIBKAT | 2020
Developing generic hydrodynamic models using artificial neural networks
| British Library Online Contents | 2002
Optimizing Neural Networks for Efficient FPGA Implementation: A Survey
| Online Contents | 2021
On the use of artificial neural networks as generic descriptors
| British Library Conference Proceedings | 1996
The use of deep neural networks for developing generic pavement rutting predictive models
| Taylor & Francis Verlag | 2022