Low Power FPGA Hardware for Machine Learning

3 Credits.

This course provides comprehensive coverage of both practical and theoretical aspects essential for designing digital systems with high speed and energy efficiency, with a specific focus on machine learning. The emphasis is placed on implementing designs for reconfigurable architectures like FPGA and conducting real-world testing of machine learning systems using an FPGA development board. Various topics will be covered, including hardware architectures, fixed-point implementation, pipelining, optimized synthesis, and routing techniques aimed at enhancing performance while reducing hardware size and power consumption. The course consists of four homework and concludes with a final project that requires hardware design using Verilog, along with evaluation through simulation and FPGA hardware. Tools to be used: Xilinx Vivado, FPGAs: Artix FPGA