I was introduced to Dieter Kranzlmüler earlier by our mutual friend and colleague, Wolfgang Gentzsch, and he graciously agreed to provide a summary of what happened at the International Symposium on High Performance Distributed Computing held in Garching earlier this month. With my thanks, here is what Dr. Kranzlmüler had to say about that conference.
170 scientists from more than 20 countries participated in the International ACM Symposium on High Performance Distributed Computing (HPDC 2009) at the Leibniz Supercomputing Centre (LRZ) of the Bavarian Academy of Science in Garching near Munich during 11 — 13 June 2009. The current scientific developments in parallel and distributed computing, high-speed networks and related software solutions have been discussed by renowned international experts.
Recognized as a scientific symposium of excellence, HPDC offers a forum which enables scientists to present their new ideas and latest technical findings, users to report challenges in locally distributed scalable applications running on high performance computers, and for technology experts to show future developments. Distributed high performance computing offers the necessary computing power, storage capacity and communication bandwidth to applications with high requirements on resources in a cost efficient way.
The HPDC conference, organized and chaired by Dieter Kranzlmüller (Ludwig-Maximilians-Universität München and Member of the Board of LRZ) was opened by Arndt Bode (Technische Universität München and Chairman of the Board of LRZ). Iannis Sagias from the European Commission introduced the main emphasis of the current EU programs.
From the 102 submitted contributions, 20 papers were selected by international experts. Three keynote speeches were delivered by Jack Dongarra, Wolfgang Gentzsch and Hans Zima. In addition, software presentations and a poster session were co-located, and a panel discussion on “How to solve the power wall problem of supercomputing (in 2015)?” was held. Supercomputer producers and sponsors Cray, Hewlett-Packard, IBM, Intel, Megware, Platform Computing und SGI, used the opportunity to present their current developments in a vendor session. The media sponsor of the event was HPCWire.
On June 9 and 10 the symposium was preceded by workshops which offered the opportunity to delve into various HPDC topics. During the CLADE workshop the challenges of large systems have been discussed, whereas MLA took a close look at monitoring, logging and account in production. The Resilience workshop discussed fault tolerance, a problem occurring particularly in high scalable systems. Further topics such as high throughput computing, benchmarking systems and applications, content management as well as quasi-opportunistic super computing were also discussed.
Contact: Prof. Dr. Dieter Kranzlmüller, Leibniz Supercomputing Centre (LRZ) of the Bavarian Academy of Science, Boltzmannstr. 1, 85748 Garching, e-mail: [email protected]
To round things out, here is a little about the LRZ
The Leibniz Supercomputing Centre (LRZ) of the Bavarian Academy of Science located on the university campus in Garching acts as an IT service provider for the universities of Munich. The LRZ supplies a high performance communication infrastructure (Münchner Wissenschaftsnetz, MWN), and offers numerous data archiving and backup mechanisms. In addition, the LRZ is the national supercomputing center and part of the Gauss Centre for Supercomputing spanning the three high performance supercomputing centers in Garching, Jülich and Stuttgart.