SFATTI: Spiking FPGA Accelerator for Temporal Task-driven Inference -- A Case Study on MNIST
By: Alessio Caviglia , Filippo Marostica , Alessio Carpegna and more
Potential Business Impact:
Makes computers recognize handwriting faster and cheaper.
Hardware accelerators are essential for achieving low-latency, energy-efficient inference in edge applications like image recognition. Spiking Neural Networks (SNNs) are particularly promising due to their event-driven and temporally sparse nature, making them well-suited for low-power Field Programmable Gate Array (FPGA)-based deployment. This paper explores using the open-source Spiker+ framework to generate optimized SNNs accelerators for handwritten digit recognition on the MNIST dataset. Spiker+ enables high-level specification of network topologies, neuron models, and quantization, automatically generating deployable HDL. We evaluate multiple configurations and analyze trade-offs relevant to edge computing constraints.
Similar Papers
A Robust, Open-Source Framework for Spiking Neural Networks on Low-End FPGAs
Neural and Evolutionary Computing
Makes brain-like computers run faster on cheap chips.
Implementation of high-efficiency, lightweight residual spiking neural network processor based on field-programmable gate arrays
Neural and Evolutionary Computing
Makes AI chips use less power for faster thinking.
FeNN: A RISC-V vector processor for Spiking Neural Network acceleration
Neural and Evolutionary Computing
Makes AI use much less power.