Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
In-memory computing devices, such as resistive RAMs, natively implement material implication or a variant of the majority-of-three operation called RM3. This operation generalizes material implication and has been used as target operation in several logic synthesis algorithms for in-memory computing applications. In this work, we investigate a homogeneous logic network data structure that uses RM3 as only logic operation. Such a data structure makes an ideal fit for the use in design automation algorithms for in-memory computing. We show how to derive RM3 networks from well-known logic synthesis data structures and a technique how to obtain such networks using technology mapping.
Giovanni De Micheli, Alessandro Tempia Calvino