Spiking Neural Network Training on Field Proarammable Gates Arrays

3 credits

This graduate reading course provides an in-depth study of training methodologies, learning rules, and hardware deployment strategies for Spiking Neural Networks (SNNs), with a focus on implementation and acceleration using Field-Programmable Gate Arrays (FPGAs). Students will critically analyze key literature in SNN learning algorithms, surrogate-gradient training, online/unsupervised learning, neuromorphic architectures, and low-power hardware design. The course emphasizes synthesizing state-of-the-art research with practical methodology for mapping SNN models onto FPGA platforms, including quantization, event-driven computation, memory architecture, and hardware-software co-design. The student will produce written reviews, technical analyses, and implementation-focused deliverables that support the supervisor's ongoing research program in hardware-accelerated neural computation and next-generation neuromorphic Al systems.