Why I like ISC...

"ISC has evolved over 20 years to the only European key event addressing all aspects of High Performance Computing. It is strongly recommended for executives from research, academia and industry who want mangement-style updates in a very professional and compact but also entertaining way."
Dr. Frank Baetke, Global HPC-Technology Program Manager, Hewlett-Packard

Why I like ISC...

"ISC has evolved over 20 years to the only European key event addressing all aspects of High Performance Computing. It is strongly recommended for executives from research, academia and industry who want mangement-style updates in a very professional and compact but also entertaining way."
Dr. Frank Baetke, Global HPC-Technology Program Manager, Hewlett-Packard

Why I like ISC...

"HPC practitioners in academia, government and industry are harnessing advanced computing to push the envelope of science and technology. Every year, ISC brings this dedicated community together to share the insights, celebrate the accomplishments and map out the future."
Kyril Faenov, General Manager HPC, Microsoft Corporation

Why I like ISC...

"As a national supercomputing center, we need to update ourselves with recent HPC technology and applications. ISC is the best place for this purpose because every year ISC invites excellent speakers and combines exciting topics and exhibition. It is also reasonable size conference, I love it."
Dr. Sik Lee, Team Leader, KISTI Supercomputing Center, Korea

Why I like ISC...

"Europe is not the continent to produce the majority of supercomputers - its contribution focusses on super algorithms and applications. ISC is the place where both fields merge in Europe and consequently a must for the annual agenda."
Prof. Dr. Frank Behrendt, Berlin Institute of Technology, Department of Energy Engineering

Why I like ISC...

"ISC provides an ideal mix of focused technical presentations, exhibitor demonstrations, and time to mingle with other participants. Attendance is small enough to have a discussion and diverse enough to get a variety of viewpoints, especially those of the key international HPC players."
Dona Crawford, Associate Director Computation, Lawrence Livermore National Laboratory

Why I like ISC...

"I have been attending ISC since 1987, I remember that being an exciting time for high-performance computing as it is today. The ISC conference has, since its beginning, been a venue for a stimulating program with the leading experts in the HPC area."
Dr. Jack Dongarra, Distinguished Professor, University of Tennessee

Why I like ISC...

"ISC offers a platform on which users from science, industry as well as vendors and industry observers have a chance to meet and discuss, unlike almost any other place in the world. As an industry representative, I have gained tremendous value from attending the event."
Dave Turek, Vice President, Deep Computing, IBM

Why I like ISC...

"I often find that I learn more at ISC compared to IEEE/ACM SCxx. The combination of more concentrated technical focus, combined with the great hospitality of the organizers, creates a unique experience. You've got to be there to really understand the draw of the event."
Dr. Horst Simon, Associate Laboratory Director, Lawrence Berkeley National Laboratory

Why I like ISC...

"I like ISC because this is where HPC Europe meets the world. ISC is the place to discuss cross-Atlantic issues and keep in touch with Asia. I like its unique atmosphere of work and recreation. ISC is where scientists meet to enjoy discussions and build communities."
Prof. Dr.-Ing. Michael Resch, Director, HLRS Stuttgart

Why I like ISC...

"ISC is the most important HPC event in Europe. ISC combines scientific and technical issues in an unique way, and makes this event most productive and enjoyable."
Dr. Andreas Adelmann, Staff Scientist, Paul Scherrer Institute (PSI)

Why I like ISC...

"I attended ISC'07 for the first time and it was a good decision. As a developing center, CHPC benefited greatly from the Hot-Seat session, where technology manufacturers were grilled. This session makes it easy for new entrants into HPC to understand the dynamics and technicalities on deciding on technologies to implement."
Dr. Happy Sithole, Director, CHPC, South Africa

Why I like ISC...

"For Cray, the ISC conference in June is a great event. ISC is a 'must attend' event, with an atmosphere more intimate and conducive to useful dialogues than larger HPC conferences in the U.S. ISC combines excellent organization, attractive suroundings, an exciting agenda and an uncompromised commitment to HPC."
Dr. Ulla Thiel, Vice President Europe, Cray

Why I like ISC...

"ISC has always been my favorite. The continuously highest international quality, the great selection of world-class speakers, the ideal mix of high-level research, business and industry talks, and Hans Meuer's unique gift to maintain a familiar atmosphere make this a unique conference and a 'must participate'!"
Prof. Dr. Wolfgang Gentzsch, EU Project DEISA

Why I like ISC...

"ISC has established itself as the key European HPC event while maintaining a technically focused atmosphere. We look forward to Intel's continued association with ISC for many years to come."
Stephan Gillich, HPC Marketing Manager EMEA, Intel

IX. HPC Challenges & Opportunities in the Petaflop/s Era

Friday, June 20, 2008
11:00 am - 12:45 pm
Lecture Hall, Level 1


ABOUT THIS SESSION

Chair: Prof. Dr. Thomas Sterling, Louisiana State University, USA

ISC’08 will conclude with an exciting session on the potential futures of our field as we look forward in to the next decade and the dawning of the Petaflops Epoch. As performance is pushed to ever greater heights continuing the trend of the last three decades new challenges emerge to inhibit further growth demanding innovative technologies and strategies to exploit them. Power consumption and cooling, floor space and cost, programmability and reliability are all now critical challenges to achieving further progress even as the community considers the possibilities for Exascale capability towards the end of the next decade. Three important presentations by world leaders in the field will demonstrate significant alternative or possibly complementing directions that may imply dramatic corner turning of the field of HPC beyond a Petaflops.

The Keynote address for this last day and session will be given by John Salmon of D.E. Shaw, a multi-time Gordon Bell Prize winner and world expert in massive N-body simulations of the large scale structure of the universe. John will provide one of this community’s first glimpses of the new domain specific architecture, Anton, which will push the boundaries of molecular dynamics simulations of proteins at the atomic level to unprecedented scale.

In our first presentation, one of the great debates of our discipline continues. Throughout the last decade the shortest distances of computing inside a conventional supercomputer has competed with the largest distances of computing spanning continents in the form of Grids. With wide diversity of problem structures and access patterns some classes of workloads have benefited from highly distributed computing resources to exceeding the capability of the world’s most powerful supercomputers. Wolfgang Gentzsch and Dieter Kranzlmüller in a manner representative of their topic will jointly present the core issues both driving and inhibiting the development and exploitation of Grid computing. Dr. Kranzlmüller is professor of Computer Science at Johannes Kepler University Linz and Project Director of the European Grid Initiative Design Study. Dr. Gentzsch of Duke University is coordinator of the German D-Grid Initiative and serves as a member of the US President’s Council of Advisors for Science and Technology.

Peter Kogge, Assistant Dean for R&D at the University of Notre Dame and an IBM Fellow (retired) has led the first major cross-discipline study of Exascale computing technologies, architectures, and systems involving a blue ribbon panel of experts from academia, industry, and government within the US. Peter will present the driving trends and factors that will characterize computing system development over the next decade across the pan-Petaflops performance regime emphasizing the strategic barriers and most promising innovations being explored.

This session will give an exciting new perspective on the future of supercomputing and its continuing potential to drive science and technology in to the 21st Century.

 

SPEAKERS IN THIS SESSION


Sponsors

The ongoing success of the International Supercomputer Conference is due in large part to our sponsors. We are pleased to highlight that Intel is the premier sponsor for ISC'08. The list of co-sponsors includes many of the major players in the HPC world. » more

Premier Sponsor

0

Co-Sponsors