We introduce functional encoding for structured connectivity such as the connectivity in convolutional layers, this offers a 58% reduction in the energy to implement a backward pass and weight update in convolutional layers when compared to existing index-based solutions. When the synaptic parameters for a 2-layer SNN trained to retain a spatio-temporal pattern, are stored using three encoding schemes: compressed row storage (PB-CSR), bitmap (PB-BMP), and crossbar (CB), we observe that PB-BMP can encode a sparser network with a 1.37x improvement in energy efficiency while suffering from a 4% increase in cost as measured by the van Rossum distance.