Figure 7
Experimental validation of DRM. Panels (a–c), (d–f) and (g–i) correspond to three separate experiments to measure the DRM. (Left column) Experimentally measured \(V_{\text {cell}}\) and the corresponding dissipated electrical power, P. Each square-shaped high–current write pulse is interleaved with a below threshold triangular read pulse (see Methods for more details). Between experiments, the key difference is the number and duration of the write pulses. (Center column) The \(I-V_{\text {cell}}\) characteristics corresponding to the read pulses are shown. The effective amorphous thickness, \(u_{\text {a}}\), is assumed to be directly proportional to the slope of these curves as per Eq. (7). (Right column) Based on the power signals shown on the left column, the \(u_{\text {a}}\) estimates based on the \(I-V_{\text {cell}}\) characteristics shown in the center column and the memristive state equations, experimental DRMs are obtained (blue traces). Also shown are the DRMs based on the steady-state values of the power signals (red traces).
