1 paper
Eric Hallman, Ilse C. F. Ipsen
We analyse the forward error in the floating point summation of real numbers, from algorithms that do not require recourse to higher precision or better hardware. We derive informa…