“…Also, circuit designers can propose variation-tolerant or variation-compensated techniques to address the variation issue ( Athmanathan et al, 2016 ; Park et al, 2017 ; Hwang et al, 2010 ; Bae et al, 2018 ), or sneak-current cancellation scheme for the sneak current issue ( Vontobel et al, 2009 ; Shevgoor et al, 2015 ; Bae et al, 2016 ). In addition, looking further forward, RRAM is regarded to be a promising candidate for in-memory computing or neuromorphic computing, because of its capability to store analog weights ( Alibart, Zamanidoost & Strukov, 2013 ; Prezioso et al, 2015 ; Yoo, 2019 ; Xue et al, 2019 ; Kim & Williams, 2019 ; Yoon, Han & Bae, 2020 ; Wang et al, 2019 ). These approaches are believed to overcome the limitation of the current computer architecture, where we need tons of inter-disciplinary research opportunity to realize them.…”