CPFloat: A C Library for Simulating Low-Precision Arithmetic

See paper: https://dl.acm.org/doi/10.1145/3585515