The goal of this Master’s thesis is to explore the power law scaling (Cherti et al. 2023) of neuromorphic algorithms and hardware by replacing conventional global backpropagation with biologically inspired local learning rules.