 |
|
|
|
| HAL : hal-00021443, version 1 |
| arXiv : cs.AR/0603115 |
| Fiche détaillée |  Récupérer au format |
|
|
|
|
| Implementation of float-float operators on graphics hardware |
|
|
| Guillaume Da Graçca 1David Defour 1 |
|
|
| (21/03/2006) |
|
|
|
| The Graphic Processing Unit (GPU) has evolved into a powerful and flexible processor. The latest graphic processors provide fully programmable vertex and pixel processing units that support vector operations up to single floating-point precision. This computational power is now being used for general-purpose computations. However, some applications require higher precision than single precision. This paper describes the emulation of a 44-bit floating-point number format and its corresponding operations. An implementation is presented along with performance and accuracy results. |
|
|
|
|
|
|
|
|
|
|
|
| 1 : | Laboratoire de Physique Appliquée et d'Automatique (LP2A) |
|
|
|
EA3679 – Université de Perpignan |
|
|
|
|
|
|
|
|
|
| Domaine | : | Informatique/Architecture Informatique/Synthèse d'image et réalité virtuelle |
|
|
| GPGPU – IEEE-754 – floating-point arithmetic – accuracy |
|
|
|
| Liste des fichiers attachés à ce document : | ||||||||||
|
||||||||||
|
|
|
| hal-00021443, version 1 | |
| http://hal.archives-ouvertes.fr/hal-00021443 | |
| oai:hal.archives-ouvertes.fr:hal-00021443 | |
| Contributeur : David Defour | |
| Soumis le : Mercredi 29 Mars 2006, 12:57:53 | |
| Dernière modification le : Mercredi 29 Mars 2006, 13:48:42 | |
