Optimal bounds for round to nearest
Optimal error bounds for rounding to nearest of arithmetic operations.
From:
Claude-Pierre Jeannerod, Siegfried M. Rump:
On relative errors of floating-point operations: Optimal bounds and applications,
Math. Comput., 87(310):803-819, 2018.
The paper also contain a specific result for division in radix2 (only the generic radix result is proved here).