Abstract: The floating-point (FP) format offers superior precision compared to its integer counterpart for multiply-and-accumulate (MAC) computations. With the rapid progress of artificial ...