Computational methods can accelerate materials science research. Density Functional Theory (DFT) modeling and machine learning (ML) are effective approaches for predicting the properties of materials absent empirical study at greatly reduced cost. Advances in semiconductor technology are contingent on the development of novel materials or novel applications of known materials. Presented are research on perovskite transition metal oxides (PV-TMO) for memristor applications, research on hydrogen-ion based memristors and research on the conduction properties of intermetallics for interconnects applications. Dynamically-resistive switching devices, memristors, form the theoretical basis for physical neural computing architecture that can overcome the data bottleneck of Von Neumann computing. PV-TMO materials that have a stable oxygen-reduced brownmillerite phase can potentially satisfy the requirements for memristors devices. Memristors based on perovskites take advantage of mixed ion and electron conductivity and switchable carrier type concentration to realize multiple stable resistive states. One study uses DFT to model the behavior of SrXLa1-XFeO3 to identify optimal compositions for such applications. Hydrogen-ion based memristors have smaller carriers, and thus allow faster migration over slower, oxygen-based alternatives. Another study uses DFT to find a metal-to-insulating transition from the introduction of hydrogen in SmNiO3, allowing its use in memristors. A growing limitation of copper metallization in advanced integrated circuits is resistivity increase affecting RC delay. Intermetallics show some promise for replacing copper. The objective study will be to utilize a dataset for the DFT modeled resistivity x mean-free path product for ML model training in the expedited screening of intermetallics as replacements for copper interconnects.