- aktuell - lists.lrz.de

Netzwartung am 7.5.
by LRZ aktuell May 6, 2013

May 6, 2013

1. Zentrale Netzdienste Der Umzug eines der beiden Zentralswitches wird morgen abgeschlossen. Zwischen 7:30 Uhr und 9:00 Uhr werden die einzelnen Serverswitches auf den neuen Zentralswitch umgehaengt. Da diese jeweils auch eine Anbindung am ersten Zentralswitch haben (Redundanz) wird es nur bei nicht vorhersehbaren Stoerungen zu Unterbrechungen kommen. Der Client-Zugang zum Internet ist von diesen Arbeiten nicht betroffen. 2. Virtuelle Firewalls Zwischen 7:35 Uhr und 8:00 Uhr wird auf den Firewall-Servicemodulen ein Softwareupdate durchgefuehrt. Dadurch wird fuer alle Institutsnetze hinter einer virtuellen Firewall fuer einige Minuten die Aussenanbindung ausfallen. 3. TUM Gabelsberger Strasse 43 Durch die Migration auf eine neue Verkabelung wird es zwischen 7:30 Uhr und 9:00 Uhr zu einer Unterbrechung kommen. 4. TUM Gabelsberger Strasse 39 Die TUM hat hier eine Stromabschaltung fuer den Zeitraum 7:30 Uhr bis 8:30 Uhr angekuendigt. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/netz/aktuell/ali4574/ Wolfgang Beyer

1 0

Call for proposals: SuperMUC Extreme Scaling Workshop
by LRZ aktuell May 2, 2013

May 2, 2013

Call for Proposals: SuperMUC Extreme Scaling Workshop Leibniz Rechenzentrum Garching July 9-11, 2013 Deadline for proposal submission: June 1st, 2013 In July 2013 LRZ is going to organize a three day workshop on extreme code scaling. The goal of this workshop is to enable scaling of user implemented software on SuperMUC, the 3 PetaFLOP/s machine at LRZ. Users can submit their own software as candidates for the workshop, and extreme scaling experts from IBM, Intel and LRZ will be working together with the users to run the software on up to 130,000 cores of SuperMUC, trying to break the magic limit of one Petaflop/s application performance. From the applicants, the most promising software packages will be selected. The main requirement to participate is: Your code already scales up to 4,000 Cores! Users with already existing LRZ or PRACE accounts should use their user account to generate the necessary scaling plots up to 4,000 cores. New users should contact Ferdinand Jamitzky (Jamitzky(a)lrz.de) for more information on how to obtain a test account. Additionally to the scaling plot, please send us the following information: Contact Data: Name, Affiliation, e-mail, phone Name of Software and short description including scientific motivation References Assurance that the code already scales up to 4000 cores (scaling plot, can also be from runs on a machine other than SuperMUC). Time needed for a single test run (< 0.5 h) Memory usage per core (< 1.6 GB) Hard disk data usage (< 1 TB) Other code requirements (libraries, I/O, compiler versions, etc) After submitting a proposal, the participants will obtain a test account with a total time budget of 50,000 cpu-hours for scaling tests and have to show that their code scales up to 4 islands on SuperMUC (1 Island consists of 8192 Intel Xeon Sandy Bridge cores). They have to present a scaling plot that shows the scaling of the computing time versus number of cores to up to 32000 cores. From the results the participants show during the test runs, 16 projects will be selected for additional runs on the whole available system during the workshop. The selected candidates will also have the opportunity to present their results during a Mini-Symposium at the ParCo2013 Conference in September in Garching. At the same time the PRACE PATC Course: Introduction to SuperMUC- the new Petaflop Supercomputer at LRZ takes place at LRZ and you are welcome to register for both. Important Dates: Proposal submission deadline: June 1, 2013 Machine Access for preparations of scaling plot: latest June 5, 2013 Test accounts will be given immediately after submission of the project proposal Scaling Plots for up to 32,000 cores: July 1, 2013 Workshop: July 9-11, 2013 This information is also available on our web server http://www.lrz-muenchen.de/aktuell/ali4573/ Ferdinand Jamitzky

1 0

Internationaler Museumstag am LRZ
by LRZ aktuell April 29, 2013

April 29, 2013

Das Museum fuer Abguesse Klassischer Bildwerke (MFA) und das Zentrum fuer Virtuelle Realitaet und Visualisierung (V2C) des Leibniz-Rechenzentrums (LRZ) beteiligen sich in diesem Jahr gemeinsam am Internationalen Museumstag. In juengster Vergangenheit haben sich einige Kooperationen zwischen verschiedensten Forschungseinrichtungen aus unterschiedlichen Themengebieten der Archaeologie und dem LRZ, mit der Absicht stereoskopische Visualisierung zu nutzen, ergeben. Mit seiner 5-seitigen Projektionsinstallation und der grossformatigen hochaufloesenden Projektionswand bietet das V2C des LRZ ideale Moeglichkeiten zur 3D-Darstellung und Bearbeitung von Datensaetzen. Beim Museumstag moechten wir der Oeffentlichkeit erste Ergebnisse dieser Verbindung modernster Technologie mit Archaeologie naeher bringen. Zum einen werden Vortraege eines internationalen Expertenteams die archaeologischen Hintergruende sowie die verwendeten Technologien erlaeutern. Zum anderen bekommen die Besucher die Moeglichkeit, ausgewaehlte Ergebnisse im Rahmen von Live-Demonstrationen im kuerzlich eroeffneten V2C zu erleben. Hierbei kommen modernste Projektionsinstallationen zum Einsatz, die das Eintauchen in virtuelle Welten erlauben. Themenschwerpunkte der Vortraege und Praesentationen sind die Digitalisierung eines Terrakottakriegers, der Vergleich und die Analyse von Statuen und Abguessen, die Dokumentation und Visualisierung eines Roemerkastells, sowie die virtuelle Darstellung einer Grabkammer in Karaburun sowie deren Rekonstruktion. Zum Abschluss der Veranstaltung wird eine Liveverbindung zur University of Tokyo aufgebaut, bei der die gemeinsame Zusammenarbeit in virtuellen Welten demonstriert wird. Die Einladung mit Abbildungen koennen Sie auf unserer Webseite als PDF-Dokument herunterladen: http://www.lrz.de/services/termine/museumstag/ * Termin: 12. Mai 2013, 14-19 Uhr * Ort: Leibniz-Rechenzentrum, Garching bei Muenchen, Boltzmannstr. 1 Um Anmeldung per E-Mail an museumstag(a)lrz.de wird gebeten. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/aktuell/ali4572/ Ludger Palm

1 0

Upd:Netzwartung am 30.4.
by LRZ aktuell April 29, 2013

April 29, 2013

Aenderungen an der Mitteilung: Tausch Switchmodul in ABK 1. RZ-Netz Zwischen 07:30 Uhr und 09:00 Uhr wird ein neuer zentraler Switch in die Rechenzentrums-Netzstruktur (Spanning Tree-Verbund) integriert. Dabei kann es (muss aber nicht) fuer die am LRZ erbrachten Dienste (Radius, E-Mail, WWW, Speicherdienste, Zugang zu Onlinemedien usw.) zu kurzzeitigen (ca. 30 Sekunden langen) Unterbrechungen kommen. Am kommenden Dienstag (7.5.) werden die Switches an denen die Server angeschlossen sind auf den neuen Zentralswitch umgehaengt. Dabei kann es wieder zu Stoerungen kommen. Der Internet-Zugang ist von diesen Arbeiten nicht betroffen. 2. Akademie der Bildenden Kuenste Zwischen 8:45 Uhr und 9:00 Uhr wird am Switch swz2-kyj ein Modul getauscht. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/netz/aktuell/ali4570/ Wolfgang Beyer

1 0

Netzwartung am 30.4.
by LRZ aktuell April 29, 2013

April 29, 2013

Zwischen 07:30 Uhr und 09:00 Uhr wird ein neuer zentraler Switch in die Rechenzentrums-Netzstruktur (Spanning Tree-Verbund) integriert. Dabei kann es (muss aber nicht) fuer die am LRZ erbrachten Dienste (Radius, E-Mail, WWW, Speicherdienste, Zugang zu Onlinemedien usw.) zu kurzzeitigen (ca. 30 Sekunden langen) Unterbrechungen kommen. Am kommenden Dienstag (7.5.) werden die Switches an denen die Server angeschlossen sind auf den neuen Zentralswitch umgehaengt. Dabei kann es wieder zu Stoerungen kommen. Der Internet-Zugang ist von diesen Arbeiten nicht betroffen. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/netz/aktuell/ali4570/ Wolfgang Beyer

1 0

Upd:Lecture Prof. Michael Flynn 'Dataflow Engines in High Performance Computing'
by LRZ aktuell April 25, 2013

April 25, 2013

Changes to this message: Update Lecture Michael Flynn: Dataflow Engines in High Performance Computing Friday, May 3rd, 2013, 10 am LRZ Hoersaal H.E.009 Abstract: Achieving parallel processor speedup (n times speedup with n processing elements) has been difficult for a broad class of applications. The problem is not technology but programming models. One answer to this speedup problem is to create an idealized data flow machine that exactly corresponds to the application and stream data through the resulting machine. In the dataflow paradigm an application is considered as a dataflow graph of the executable actions.. The extraordinary density achieved by FPGAs allows emulations of the application dataflow machine, providing more than an order of magnitude speedup even as executed as an emulation of the data flow machine. Michael Flynn received his Ph.D. from Purdue University in 1961. He joined IBM in 1955 and for ten years worked in the areas of computer organization and design. He was design manager of prototype versions of the IBM 7090 and 7094/II, and later for the System 360 Model 91 Central Processing Unit. Between 1966 and 1974 Prof. Flynn was a faculty member of Northwestern University and the Johns Hopkins University. In 1975 he became Professor of Electrical Engineering at Stanford University, and was Director of the Computer Systems Laboratory from 1977 to 1983. He was founding chairman of both the ACM Special Interest Group on Computer Architecture and the IEEE Computer Society's Technical Committee on Computer Architecture. Prof. Flynn was the 1992 recipient of the ACM/IEEE Eckert-Mauchley Award for his technical contributions to computer and digital systems architecture. He was the 1995 recipient of the IEEE-CS Harry Goode Memorial Award in recognition of his outstanding contribution to the design and classification of computer architecture. In 1998 he received the Tesla Medal from the International Tesla Society (Belgrade), and an honorary Doctor of Science from Trinity College (University of Dublin), Ireland. He is the author of three books and over 250 technical papers. This information is also available on our web server http://www.lrz-muenchen.de/aktuell/ali4569/ Ludger Palm

1 0

Lecture Prof. Michael Flynn 'Dataflow Engines in High Performance Computing'
by LRZ aktuell April 25, 2013

April 25, 2013

Lecture Michael Flynn: Dataflow Engines in High Performance Computing Friday, May 3rd, 2013 LRZ Hoersaal H.E.009 Abstract: Achieving parallel processor speedup (n times speedup with n processing elements) has been difficult for a broad class of applications. The problem is not technology but programming models. One answer to this speedup problem is to create an idealized data flow machine that exactly corresponds to the application and stream data through the resulting machine. In the dataflow paradigm an application is considered as a dataflow graph of the executable actions.. The extraordinary density achieved by FPGAs allows emulations of the application dataflow machine, providing more than an order of magnitude speedup even as executed as an emulation of the data flow machine. Michael Flynn received his Ph.D. from Purdue University in 1961. He joined IBM in 1955 and for ten years worked in the areas of computer organization and design. He was design manager of prototype versions of the IBM 7090 and 7094/II, and later for the System 360 Model 91 Central Processing Unit. Between 1966 and 1974 Prof. Flynn was a faculty member of Northwestern University and the Johns Hopkins University. In 1975 he became Professor of Electrical Engineering at Stanford University, and was Director of the Computer Systems Laboratory from 1977 to 1983. He was founding chairman of both the ACM Special Interest Group on Computer Architecture and the IEEE Computer Society's Technical Committee on Computer Architecture. Prof. Flynn was the 1992 recipient of the ACM/IEEE Eckert-Mauchley Award for his technical contributions to computer and digital systems architecture. He was the 1995 recipient of the IEEE-CS Harry Goode Memorial Award in recognition of his outstanding contribution to the design and classification of computer architecture. In 1998 he received the Tesla Medal from the International Tesla Society (Belgrade), and an honorary Doctor of Science from Trinity College (University of Dublin), Ireland. He is the author of three books and over 250 technical papers. This information is also available on our web server http://www.lrz-muenchen.de/aktuell/ali4569/ Ludger Palm

1 0

Upd:SuperMUC: Returned to user operation
by LRZ aktuell April 24, 2013

April 24, 2013

Changes to this message: Returned to user operation Dear users of the SuperMUC Petaflop system, The scheduled maintenance has been completed, and the system has been returned to user operation. Login to the machine is expected to be possible at ca. 17:30 on April 24. The changes to the system are as follows: * Updated operating environment (SLES11SP2). Please consult the release notes by SuSE/Novell at https://www.suse.com/releasenotes/ x86_64/SUSE-SLES/11-SP2/for details. In particular, we expect that the Transparent Huge Pages support should give a performance improvement for many applications. Please provide feedback if you happen to observe the opposite effect for yours! Also, some applications may require significantly more memory, so again please report to us if an application fails with allocation errors that used to run fine before. * Bug fix update for the MKL libraries * Bug fix and minor feature update of the (non-default) Intel compiler version 13.1 * Bug fix update for Intel MPI 4.1 (MPI_Abort() now should really terminate the complete job). However, Intel MPI 4.0 (and earlier) should not be used under this new environment any more. Please note that we expect existing binaries to still run correctly in the new environment. Should this not be the case, please recompile your executable and resubmit your job(s) before opening an incident at our service desk. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4557/ Reinhold Bader

1 0

Upd:TSM Kernel- und Firmware Updates am 22.04.2013 Beendet
by LRZ aktuell April 22, 2013

April 22, 2013

Aenderungen an der Mitteilung: Update Update 22.04.2013 16:00 Uhr: Die Wartung ist beendet. Alle Systeme sollten nun wieder verfuegbar sein. Um Ihnen auch in Zukunft die gewohnte Servicequalitaet zur Verfuegung stellen zu koennen, fuehren wir am Montag, den 22.04.2013 von 7:00 bis ca. 17:00 Uhr eine Hard- und Softwarewartung an unserem Archiv- und Backupsystem durch. Im Zuge dieser Wartung kann es Beeintraechtigungen bei den Tapemounts und bei der Verfuegbarkeit der Server geben. Wir bitten um Ihr Verstaendnis, falls es zu Unannehmlichkeiten auf Ihrer Seite kommen sollte. In order to be able to keep delivering the familiar service quality for you we perform a hard- and software maintenance on our Archive and Backup System on Monday, 2013-04-22 from 7:00 am to 17:00. We apologize for any inconvenience on your side. Note: SuperMIG/MUC Archive Nodes, HLRB2 Archive Nodes, Grid as well as Linux-Cluster users are affected too. Ihr ABS-Team Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/datenhaltung/adsm/aktuell/ali4554/ Monika Hufnagl

1 0

Upd:SuperMIG: Returned to user operation
by LRZ aktuell April 22, 2013

April 22, 2013

Changes to this message: Update Dear users of the SuperMIG migration system at LRZ, The system has been returned to user operation. The cause for the system crash was a software problem that only occurs very rarely. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4566/ Reinhold Bader

1 0