The arti cially generated electrical signal, in retinal prostheses is very dierent from the natural neuronal signal and dicult for the brain to interpret. Computer models in general -the present one in particular- have the ability to simulate the behaviour of the neuronal network in ways such as comparing dierent sizes of stimulating electrode con guration. However, better simulation methods are needed to fully understand these neuronal mechanisms and improve their medical applications. The objective of this thesis is to identify the optimal electrode spacing of a retinal implant that can be safely used, so that patients suering from retinal degenerative diseases can pro t from better-constructed implants. For this purpose, the energy consumption of mammalian, amphibian and cephalopodal (squid) neuronal membranes have been compared. The mammalian membrane was modelled after the Fohlmeister et al 2010 model, the amphibian membrane after Fohlmeister et al 1997 model and the squid membrane after Hodgkin and Huxley. Furthermore, their detailed energy usage was analysed, including simultaneous stimulating and inhibiting properties. Especially for this thesis, a computer model has been implemented using MATLAB and the nite element software Comsol Multiphysics in order to predict the activation of retinal ganglion cells during epi-retinal stimulation. The activation, driven by the membrane model of Fohlmeister et al 2010 was simulated on dierent parts of the neuron (sodium channel band and distant axon) with various electrode spacings. The comparison between the membranes of dierent taxonomic classes showed a partially increased energy consumption in sodium and potassium channels of the higher developed mammalian membrane compared to the amphibian membrane, which was compensated by improved synchronization between the dierent ion channel types, allowing for better cross eciency. The simulation of various electrode spacings showed a dependency between the stimulating energy on the neuron and the electrode spacing. Further- more, an ideal electrode spacing was identi ed, where the neuron was able to stimulate with least amount of energy, without wasting valuable space on the electrode carrier. The chosen membrane models and activating model, together with the methods and the performed simulations, proved to be the right combination needed in order to reach the objective of this thesis. Bene ting from the knowledge of the characteristics of the nervous system and from the capacity of today's science, patients suering from retinal degeneration diseases are given the chance to have their sense of vision restored.