Deep Neural Network using Stochastic Computing
- Duration: Sep. 2017 – June 2018.
- This is the final project of the undergraduate course “Digital Circuit Lab” in NTU. I collaborated with 2 other classmates on this project.
- We implemented Multiply-Accumulate (MAC) and various operations in deep neural network via stochastic computation.
- The goal is tp build a neural network with lower hardware cost in terms of power consumption area-delay product on FPGA boards (Altera DE10-nano).
- Our team is one of the Regional Finalists in Greater China of Innovate FPGA 2018.
- Click the links to view the slides, the poster and the demo video.