Algorithm–Hardware Co-Design of Silicon Nanowire Devices for Neuromorphic Computing Applications In this thesis silicon nanowire memristive devices are investigated as a multifunctional platform for neuromorphic computing and neuromorphic biosensing. Using CMOS-compatible fabrication and oxide engineering, silicon nanowire devices with stable resistive switching behavior were developed and experimentally characterized. The results show that oxide engineering allows a single silicon nanowire platform to support different neuromorphic functions, such as synaptic weight storage, nonlinear activation, pulse generation, and neuron-like behavior. To evaluate their suitability for neuromorphic hardware, the devices were integrated into a hardware-aware simulation framework using CrossSim. Artificial neural networks (ANNs) and ternary neural networks (TNNs) were mapped onto memristive crossbar architectures using experimentally measured device characteristics.