Carregant...
Carregant...

Vés al contingut (premeu Retorn)

Exploiting asynchrony from exact forward recovery for DUE in iterative solvers

Autor
Jaulmes, L.; Casas, M.; Moreto, M.; Ayguade, E.; Labarta, J.; Valero, M.
Tipus d'activitat
Document cientificotècnic
Data
2015
Codi
UPC-DAC-RR-CAP-2015-5
Repositori
http://hdl.handle.net/2117/110733 Obrir en finestra nova
Resum
This paper presents a method to protect iterative solvers from Detected and Uncorrected Errors (DUE) relying on error detection techniques already available in commodity hardware. Detection operates at the memory page level, which enables the use of simple algorithmic redundancies to correct errors. Such redundancies would be inapplicable under coarse grain error detection, but become very powerful when the hardware is able to precisely detect errors. Relations straightforwardly extracted from t...
Citació
Jaulmes, L., Casas, M., Moreto, M., Ayguadé, E., Labarta, J., Valero, M. "Exploiting asynchrony from exact forward recovery for DUE in iterative solvers". 2015.
Paraules clau
BiCGStab, Conjugate gradient, Fault tolerance, Forward recovery, GMRES, HPC, Interpolation, Krylov subspace, Resilience
Grup de recerca
CAP - Grup de Computació d'Altes Prestacions

Participants

Arxius

  • 6.pdf  6.pdf (1478655 bytes)