Survey on FPGA-based Accelerators for CNNs

[sparsh]

CNNs (convolutional neural networks) have been recently successfully applied for a wide range of cognitive challenges. Given high computational demands of CNNs, custom hardware accelerators are vital for boosting their performance. The high energy-efficiency, computing capabilities and reconfigurability of FPGA make it a promising platform for hardware acceleration of computation-intensive CNNs.


Our paper surveys 75+ techniques for implementing and optimizing CNN algorithms on FPGA. Accepted in Neural Computing and Applications journal 2018.

PDF is attached 2018_NCAA_Mittal_FPGA_Accelerator_CNN.pdf

[Eric Esteve]

I didn't read this paper in depth, but it looks very detailed. Happy to see that researchers are using Semiwiki to share their hard work!
Just a question: what is your business model, if you share for free such a good work?

[sparsh]

Thanks a lot for your kind words. As a researcher, I am happy to share my insights and knowledge. If researchers/developers read or refer my articles, this a sufficient encouragement and reward for me.