The First International ComplexHPC Challenge

Introduction

The First International ComplexHPC Challenge is an integral part of the ComplexHPC Spring School 2011, and is open to all Spring School participants. The Challenge is an excellent opportunity to showcase your work to your peers - and even win a prize in the process. Selected Challenge participants have two opportunities to show a (preferably live) demonstration of their work: a poster and demo session on the first day of the Spring School, and an individual 20-minute presentation and demo on the final day.

Objective

The ComplexHPC project description states:

In different fields of science and engineering it is necessary to solve complex and challenging problems with high computational cost. For this purpose, scientists and engineers normally use homogeneous high performance computers. Nowadays, the emergence of heterogeneous computing allows research groups, enterprises and educational institutions to use networks of processors which are already available. On the other hand, high performance computers have become more and more hierarchical and heterogeneous (e.g., a cluster of multiprocessor nodes using multicore processors). These modern hierarchical and heterogeneous computing infrastructures are hard to program and use efficiently, particularly for extreme-scale computing. Consequently, none of the state-of-the-art solutions are able to efficiently use such environments.

The objective of the First International ComplexHPC Challenge is to demonstrate and highlight state-of-the-art problem solving in the ComplexHPC field, recognizing novel, real-world solutions to the problems present in today's complex, distributed, heterogeneous, and hierarchical computing environments. Solving these problems will realize the ultimate goal of the ComplexHPC project: easy and efficient use of any type of compute resources available to scientists and engineers.

Call for Proposals

The First International ComplexHPC Challenge will focus on the easy and efficient use of ComplexHPC resources, including (combinations of) grids, clouds, clusters, GPUs, multicores, or any other type of compute resources. Challenge participants must show how their system makes 'efficient' use of ComplexHPC resources, where efficiency can be a measure of performance, scalability, cost, energy use, etc. For use as a possible (or partial) target platform, all selected Challenge participants will be granted access to the DAS-4 system, immediately after acceptance of their proposal.

We invite Spring School participants to submit a short paper (4 pages or less; shorter is better) describing their system. A proposal should contain:

• A description of the problem solved, its significance, and relevance to the (research) community.
• A description of the system proposed.
• The architecture, and the underlying technology used.
• A description of the (heterogeneous and hierarchical) resources used.
• Quantitative data on the performance and/or scalability of the resulting system.
• A thorough analysis of the efficiency of the system.

Proposals should be send to Jason Maassen (jason AT cs.vu.nl).

Based on the proposals, a maximum of 4 Spring School participants will be invited to demonstrate their system during the ComplexHPC Spring School on May 11. The jury will select a winner based on the demonstration and the accompanying presentation. Entries will be judged on the quality of the demonstration, presentation quality, the extent in which combinations of heterogeneous and hierarchical resources are being used, and the overall efficiency of the system. Live demonstrations are strongly encouraged!

On May 9th, the participants have the opportunity to test and demonstrate their system during the poster and demo session.

Timeline

Presentation Schedule for May 11

Prizes

Contact

• Jason Maassen, VU University, Amsterdam, The Netherlands.
• Niels Drost, VU University, Amsterdam, The Netherlands.