Abstract: The utilization of reconfigurable and flexible acceleration for compute-intensive kernels, e.g., neural networks, crypto-engines, and arithmetics, have been on the rise in recent years, ...