Abstract
On a machine with addition and substraction instructions, an overflow indicator and a branch on overflow instruction, but without multiplication or division capabilities, a fast algorithm for division is given wherein a small amount of additional hardware can be used to implement the algorithm. The algorithm was programmed to implement double-precision floating-point division for IBM System/3, with considerable improvement over the conventional methods.
| Original language | English |
|---|---|
| Pages (from-to) | 2137-2138 |
| Number of pages | 2 |
| Journal | IBM technical disclosure bulletin |
| Volume | 17 |
| Issue number | 7 |
| State | Published - 1974 |