On the minimum error in addition processes of positive floating-point numbers
Volume 13 / 1973
Applicationes Mathematicae 13 (1973), 351-366
DOI: 10.4064/am-13-3-351-366