A Heterogeneous Stochastic Computing Multiplier for Universally Accurate and Energy-Efficient DNNs

Stochastic computing (SC) has become a promising approximate computing solution by its negligible resource-occupancy and ultra-low energy-consumption. As a potential replacement of accurate multiplication, SC can dramatically mitigate the problematic power consumption by DNNs. However, current SC-multipliers illustrate an extremely imbalanced accuracy across product space - nearly float-point accu...