“…Efforts offering modified march test algorithms that involve specific sequences of sequential memory (read, write) operations to optimize test time [7,8] or improve coverage [9][10][11][12] fail in the detection of unique RRAM faults [5] and typically DFT schemes are introduced to further enable improved fault coverage (FC) and/or optimized This work was supported by the EU H2020 grant "DAIS" with funding from the ECSEL Joint Undertaking under grant agreement No 101007273. test time. However, existing dedicated DFTs are expensive in terms of hardware [13], optimistic regarding variations and lack implementations [14], impractical due to large voltage requirements [6,15,16] and exhibit functional issues due to to the reliance on sneak-paths [7,17]. DFTs that target high FC typically involve slow, probabilistic write operations [6,[18][19][20][21][22].…”