Bioinspired semiconductor-based devices with adaptive and dynamic properties will have many advantages over conventional static digital silicon-based technologies. The ability to compute, process, and retain information in parallel, without referencing other circuit elements, offers enhanced speed, storage density, energy efficiency, and functionality benefits. A novel crossbar microwire-based device consisting of Nb/NbO/Pt structure that exhibits neural synapse-like adaptive conductivity (i.e., synaptic plasticity) is presented. The neuromorphic memristive junction, formed at the interface between the Pt metal wire and the thermally annealed core-shell Nb-NbO wire, demonstrates 1000 times conductivity change with an effective continuum of resistance levels. The device can also be fully activated to display standard resistance switching between two states. In the subthreshold regime, the voltage flux applied through the ∼400 nm thick NbO junction is shown to have a linear relationship to the charge produced within the device. The conductance value G is a function of the total flux history applied. The linear flux-charge relationship is exploited to demonstrate the voltage-pulse invariance. This suggests that only the integrated flux produced during a voltage-pulse application determines the charge generated within the junction, regardless of the operational parameters like voltage amplitude and time interval. Variation in current onset voltage as a function of flux is discussed with reference to carrier extraction at the intrinsically doped metal-semiconductor interface. The observed flux invariance has implications in emerging neuromorphic semiconductor hardware. Enabling pulse stimuli to be designed to have equal flux through the device from nonvoltage sources such as light may increase functionality in bioinspired computing and applications.