Scalable Hierarchical Algorithms for eXtreme Computing Workshop

Apr 28 2012 08:00 AM - Apr 30 2012 04:00 PM

​The 2012 SHAX-C workshop focuses international expert attention on the prospects for the three great hierarchical algorithms of scientific computing: multigrid, fast transforms, and fast multipole methods. These methods are kernels in simulations based on formulations of partial differential equations, integral equations, and interacting particles – in short, they are scientific and engineering workhorses. 

There is encouraging early evidence that fast multipole methods and their treecode cousins fare relatively well on extreme architectures. It is therefore opportune to bring together leading developers and practitioners of hierarchical algorithms of diverse stripes in a workshop aimed at sharing knowledge, practice, opinions about architecture targets, and perhaps even code. The workshop is sponsored in part by KAUST's office of competitive research funding.

The workshop duration will be three working days (April 28-30), preceded by an optional Saudi weekend social day, in which guest participants can adjust to the time difference while enjoying KAUST’s resort environment with regular on-campus participants. The working days of the conference are Saturday-Monday, with the pre-workshop day a Friday. The first day will feature algorithms for each of the three hierarchical families, the second day freely available software that scales to contemporary limits, including analysis of per-node performance and scalability, with an evening poster session. The third day will be devoted to cross-fertilization and formulation of directions for algorithms, programming models, and hardware.​

Organizing Committee

  • Lorena Barba, Assistant Professor, Boston University
  • David Keyes, Professor, Dean of the MCSE Division, Director of the Center for Extreme Computing, KAUST
  • Matthew Knepley, Senior Research Associate, University of Chicago
  • Hatem Ltaief, Computational Scientist, Supercomputing Laboratory, KAUST
  • Rio Yokota, Research Scientist, Center of Extreme Computing, KAUST​

Papers recommened by participants

Preparing Algebraic Multigrid for Exascale​

Shared by Ulrike Yang

Faster Topology-aware Collective Algorithms Through Non-minimal Communication

Shared by Jack Poulson

Parallelizing a Black-Scholes Solver Based on Finite Elements and Sparse Grids

Shared by Dirk Pflüger

Multi- and Many-Core Data Mining with Adaptive Sparse Grids​
Shared by Dirk Pflüger

Hierarchical QR factorization algorithms for multi-core cluster systems
Shared by Julien Langou

Tiled QR factorization algorithms​
Shared by Julien Langou

On the Communication Complexity of 3D FFTs and its Implications for Exascale​

Shared by Rich Vuduc

Parallel Black Box H-LU Preconditioning for Elliptic Boundary Value Problems

Shared by Sabine LeBorne

Secure web-based communication​
Shared by Nighat Mir

A Cartesian treecode for screened coulomb interactions​
Shared by Hans Johnston