HISQ inverter on Intel Xeon Phi and NVIDIA GPUs
Mukherjee S, Kaczmarek O, Schmidt C, Steinbrecher P, Wagner M (2015)
In: Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014. PoS: 044.
Konferenzbeitrag
| Veröffentlicht | Englisch
Download
Es wurden keine Dateien hochgeladen. Nur Publikationsnachweis!
Autor*in
Mukherjee, Swagato;
Kaczmarek, OlafUniBi 
;
Schmidt, ChristianUniBi ;
Steinbrecher, PatrickUniBi;
Wagner, Mathias
;
Schmidt, ChristianUniBi ;
Steinbrecher, PatrickUniBi;
Wagner, MathiasAbstract / Bemerkung
The runtime of a Lattice QCD simulation is dominated by a small kernel, whichcalculates the product of a vector by a sparse matrix known as the "Dslash"operator. Therefore, this kernel is frequently optimized for various HPCarchitectures. In this contribution we compare the performance of the IntelXeon Phi to current Kepler-based NVIDIA Tesla GPUs running a conjugate gradientsolver. By exposing more parallelism to the accelerator through invertingmultiple vectors at the same time we obtain a performance 250 GFlop/s on botharchitectures. This more than doubles the performance of the inversions. Wegive a short overview of both architectures, discuss some details of theimplementation and the effort required to obtain the achieved performance.
Erscheinungsjahr
2015
Titel des Konferenzbandes
Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014
Seite(n)
044
Konferenz
32nd International Symposium on Lattice Field Theory (Lattice 2014)
Konferenzort
Brookhaven, NY, USA
Konferenzdatum
2014-06-23 – 2014-06-28
Page URI
https://pub.uni-bielefeld.de/record/2765240
Zitieren
Mukherjee S, Kaczmarek O, Schmidt C, Steinbrecher P, Wagner M. HISQ inverter on Intel Xeon Phi and NVIDIA GPUs. In: Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014. PoS; 2015: 044.
Mukherjee, S., Kaczmarek, O., Schmidt, C., Steinbrecher, P., & Wagner, M. (2015). HISQ inverter on Intel Xeon Phi and NVIDIA GPUs. Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014, 044. PoS.
Mukherjee, Swagato, Kaczmarek, Olaf, Schmidt, Christian, Steinbrecher, Patrick, and Wagner, Mathias. 2015. “HISQ inverter on Intel Xeon Phi and NVIDIA GPUs”. In Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014, 044. PoS.
Mukherjee, S., Kaczmarek, O., Schmidt, C., Steinbrecher, P., and Wagner, M. (2015). “HISQ inverter on Intel Xeon Phi and NVIDIA GPUs” in Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014 (PoS), 044.
Mukherjee, S., et al., 2015. HISQ inverter on Intel Xeon Phi and NVIDIA GPUs. In Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014. PoS, pp. 044.
S. Mukherjee, et al., “HISQ inverter on Intel Xeon Phi and NVIDIA GPUs”, Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014, PoS, 2015, pp.044.
Mukherjee, S., Kaczmarek, O., Schmidt, C., Steinbrecher, P., Wagner, M.: HISQ inverter on Intel Xeon Phi and NVIDIA GPUs. Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014. p. 044. PoS (2015).
Mukherjee, Swagato, Kaczmarek, Olaf, Schmidt, Christian, Steinbrecher, Patrick, and Wagner, Mathias. “HISQ inverter on Intel Xeon Phi and NVIDIA GPUs”. Proceedings, 32nd International Symposium on Lattice Field Theory (Lattice 2014) : Brookhaven, NY, USA, June 23-28, 2014. PoS, 2015. 044.
Export
Markieren/ Markierung löschen
Markierte Publikationen
Quellen
arXiv: 1409.1510
Inspire: 1315118
Suchen in
Google Scholar