Neuromorphic computing systems such as DYNAPs and Loihi have recently been\nintroduced to the computing community to improve performance and energy\nefficiency of machine learning programs, especially those that are implemented\nusing Spiking Neural Network (SNN). The role of a system software for\nneuromorphic systems is to cluster a large machine learning model (e.g., with\nmany neurons and synapses) and map these clusters to the computing resources of\nthe hardware. In this work, we formulate the energy consumption of a\nneuromorphic hardware, considering the power consumed by neurons and synapses,\nand the energy consumed in communicating spikes on the interconnect. Based on\nsuch formulation, we first evaluate the role of a system software in managing\nthe energy consumption of neuromorphic systems. Next, we formulate a simple\nheuristic-based mapping approach to place the neurons and synapses onto the\ncomputing resources to reduce energy consumption. We evaluate our approach with\n10 machine learning applications and demonstrate that the proposed mapping\napproach leads to a significant reduction of energy consumption of neuromorphic\ncomputing systems.\n