<p dir="ltr">As edge computing architectures bring processing closer to data sources, there is an increasing need for memory technologies that can work effectively and consistently in a variety of situations while using minimal energy. Memristors are memory devices that have the potential to greatly increase the performance and scalability of edge devices. However, a key challenge is to achieve precise resistance switching. Silicon (Si) surfaces embedded in a proton-conducting polymer can demonstrate controllable memristor behavior wherein hydrogen (H) atoms are deposited onto the surface. When H is inside the polymer, its conductivity decreases. When H is on the silicon surface, its bulk conductivity increases due to more mid-gap traps. Migration of H atom placement can make a memristor unit cell whose impedance modulates in response to electrical signals. This study investigates the critical function of H atoms by deliberately altering their position and concentration within and upon silicon-based memristor devices. Using the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS), we investigate the impact of temperature (T) and electric field (EF) on H migration. We define a polygonal volume of Si and deposit H atoms on its top surface. After energy minimization, we apply T and EF to observe diffusion and drift of H atoms. The hopping rate depends on applied T and EF. We thus establish a relationship between the three-dimensional velocity of H and applied T and EF. We simulate several movement pathways of H atoms over time under the influence of varying T and EF acting separately or simultaneously. Therefore, we can determine the required magnitude and direction of EF and T to be introduced to the system to achieve desired H location, concentration, and configuration. Finally, we assess the device performance at different T and EF to assess memory retention rate. Our approach aims to enhance the functionality of edge computing devices and enable more effective neuromorphic computing that can emulate human brain operations. However, the limitations of this study include potential scalability issues and the necessity for precise control over hydrogen dispersion. Despite these challenges, the research provides valuable insights on how to modify the electrical characteristics of memristors, offering a way forward in the development of advanced silicon based electronic devices.</p>