The recently proposed Posit number system has been regarded as a particularly well-suited floating-point format to optimize the throughput and efficiency of low-precision computations in convolutional ...