Graphic Processing Units offer the possibility of executing several threads in parallel, providing the user with higher throughput with respect to traditional multi-core processors. However, the additional resources required to schedule and handle the parallel processes may have the countermeasure of making GPUs more prone to be corrupted by neutrons. The reported experimental results prove that un-optimized thread distribution may exacerbate the device output error rate. As demonstrated, increasing the parallel algorithm block size minimizes the GPU neutron-induced output error rate. The GPU parallelism management is then analyzed as a method to increase reliability. © 2013 IEEE.
Graphic Processing Units offer the possibility of executing several threads in parallel, providing the user with higher throughput with respect to traditional multi-core processors. However, the additional resources required to schedule and handle the parallel processes may have the countermeasure of making GPUs more prone to be corrupted by neutrons. The reported experimental results prove that un-optimized thread distribution may exacerbate the device output error rate. As demonstrated, increasing the parallel algorithm block size minimizes the GPU neutron-induced output error rate. The GPU parallelism management is then analyzed as a method to increase reliability. © 2013 IEEE.
Threads Distribution Effects on Graphics Processing Units Neutron Sensitivity / Rech, P., Fairbanks, T.D., Quinn, H.M., Carro, L.. - In: IEEE TRANSACTIONS ON NUCLEAR SCIENCE. - ISSN 0018-9499. - 2013, 60:6(2013), pp. 4220-4225. [10.1109/TNS.2013.2286970]
Threads Distribution Effects on Graphics Processing Units Neutron Sensitivity
Rech P.;
2013-01-01
Abstract
Graphic Processing Units offer the possibility of executing several threads in parallel, providing the user with higher throughput with respect to traditional multi-core processors. However, the additional resources required to schedule and handle the parallel processes may have the countermeasure of making GPUs more prone to be corrupted by neutrons. The reported experimental results prove that un-optimized thread distribution may exacerbate the device output error rate. As demonstrated, increasing the parallel algorithm block size minimizes the GPU neutron-induced output error rate. The GPU parallelism management is then analyzed as a method to increase reliability. © 2013 IEEE.| File | Dimensione | Formato | |
|---|---|---|---|
|
TNS_Threads_Distribution_Effects_on_Graphics_Processing_Units_Neutron_Sensitivity.pdf
Solo gestori archivio
Descrizione: IEEE Transactions on Nuclear Science - conference paper
Tipologia:
Versione editoriale (Publisher’s layout)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
738.33 kB
Formato
Adobe PDF
|
738.33 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione
