1 paper
James Quinlan, E. Theodore L. Omtzigt
This research investigates using a mixed-precision iterative refinement method using posit numbers instead of the standard IEEE floating-point format. The method is applied to solv…