Shadow computation with BFloat16 to compute numerical accuracy
| Autor: | David Defour, Pablo de Oliveira Castro, Matei Istoan, Eric Petit |
|---|---|
| Přispěvatelé: | LAboratoire de Mathématiques et PhySique (LAMPS), Université de Perpignan Via Domitia (UPVD), ANR-20-CE46-0009,INTERFLOP,Plateforme d'analyse pour l'arithmétique flottante(2020) |
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: | |
| Zdroj: | IEEE 28th Symposium on Computer Arithmetic (ARITH) IEEE 28th Symposium on Computer Arithmetic (ARITH), Jun 2021, Virtual Conference, France HAL |
| Popis: | International audience; In this article, we propose to exploit the new computational capability offered by the Bfloat16 representation format to perform shadow computations and compute estimations of the relative error. We demonstrate and evaluate the assumptions under which shadow computation is valid for the summation problem. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|