Spiking neural networks enabling massively parallel, low-power & low-latency computation