Tag fast2multfma

Computing the Error of a FP Addition or Multiplication

Let \(a,b\) be 2 floating point numbers. It can be shown that \((a+b)-\RN(a+b)\) is a floating point number. This may not be true for...