Hardware-based acceleration approaches for Machine Learning (ML) workloads have been embracing the significant potential of post-CMOS switching devices to attain reduced footprint and/or ...