The Hardware Impact of Quantization and Pruning for Weights in Spiking Neural Networks