Iwainsky, Christian ; Reichstein, Thomas ; Dahnken, Christopher ; Mey, Dieter an ; Terboven, Christian ; Semin, Andrey ; Bischof, Christian (2011)
An Approach to Visualize Remote Socket Traffic on the Intel Nehalem-EX.
16th International Conference on Parallel Computing (Euro-Par 2010). Ischia, Italy (31.08.2010-03.09.2010)
doi: 10.1007/978-3-642-21878-1_64
Konferenzveröffentlichung, Bibliographie
Kurzbeschreibung (Abstract)
The integration of the memory controller on the processor die enables ever larger core counts in commodity hardware shared memory systems with Non-Uniform Memory Architecture properties. Shared memory parallelization with OpenMP is an elegant and widely used approach to leverage the power of such systems. The binding of the OpenMP threads to compute cores and the corresponding memory association are becoming even more critical in order to obtain optimal performance. In this work we provide a method to measure the amount of remote socket memory accesses a thread generates. We use available performance monitoring CPU counters in combination with thread binding on a quad socket Nehalem EX system. For visualization of the collected data we use Vampir.
| Typ des Eintrags: | Konferenzveröffentlichung |
|---|---|
| Erschienen: | 2011 |
| Autor(en): | Iwainsky, Christian ; Reichstein, Thomas ; Dahnken, Christopher ; Mey, Dieter an ; Terboven, Christian ; Semin, Andrey ; Bischof, Christian |
| Art des Eintrags: | Bibliographie |
| Titel: | An Approach to Visualize Remote Socket Traffic on the Intel Nehalem-EX |
| Sprache: | Englisch |
| Publikationsjahr: | 2011 |
| Ort: | Berlin, Heidelberg |
| Verlag: | Springer |
| Buchtitel: | Euro-Par 2010 Parallel Processing Workshops |
| Reihe: | Lecture Notes in Computer Science |
| Band einer Reihe: | 6586 |
| Veranstaltungstitel: | 16th International Conference on Parallel Computing (Euro-Par 2010) |
| Veranstaltungsort: | Ischia, Italy |
| Veranstaltungsdatum: | 31.08.2010-03.09.2010 |
| DOI: | 10.1007/978-3-642-21878-1_64 |
| Kurzbeschreibung (Abstract): | The integration of the memory controller on the processor die enables ever larger core counts in commodity hardware shared memory systems with Non-Uniform Memory Architecture properties. Shared memory parallelization with OpenMP is an elegant and widely used approach to leverage the power of such systems. The binding of the OpenMP threads to compute cores and the corresponding memory association are becoming even more critical in order to obtain optimal performance. In this work we provide a method to measure the amount of remote socket memory accesses a thread generates. We use available performance monitoring CPU counters in combination with thread binding on a quad socket Nehalem EX system. For visualization of the collected data we use Vampir. |
| Fachbereich(e)/-gebiet(e): | 20 Fachbereich Informatik 20 Fachbereich Informatik > Scientific Computing Zentrale Einrichtungen |
| Hinterlegungsdatum: | 10 Jul 2012 14:15 |
| Letzte Änderung: | 04 Aug 2021 16:00 |
| PPN: | |
| Export: | |
| Suche nach Titel in: | TUfind oder in Google |
 |
Frage zum Eintrag |
Optionen (nur für Redakteure)
 |
Redaktionelle Details anzeigen |
Drucken
Impressum