Academic track

The industrial talks were followed by a more academic track the following day. These papers along with the introductory papers by Elster and her students, are included in the ParCo 2009 proceedings edited by G. Joubert et. al , published by IOS Press. PDF files for personal use only are included below .

These articles include two papers looking at GPU solvers:

• Porus Rock Simulations and Lattice Boltzmann on GPUs (PDF for personal use) by Erik Ola Aksnes and Anne C. Elster (both NTNU, Norway) which looked at using the Lattice Boltzman Method on large 3D datasets on GPUs for fluid similations, and
  

• An efficient multi-algorithms sparse linear solver for GPUs (PDF for personal use) by Stéphane Vialle; Thomas Jost and Sylvain Constassot-Vivier (all from Supé lec Campus de Metz, Franc) which discussed how to implement a sparse solver on GPUs.

The EuroGPU 2009 workshop/minisymposium was rounded off by two GPU modeling presentations:

• Abstraction of Programming Models Across Multi-Core and GPGPU Architectures (PDF for personal use) by Ian Grimstead and David R. Walker (Cardiff University, UK) , and
  

• Modeling Multi-GPU Systems (PDF for personal use) by Daniele G. Spampinato, Anne C. Elster and Thorvald Natvig (all from NTNU, Norway).

The last two papers should provide tools for those who want to get a good feel for the performance potential CPU and multi-core CPU versus GPU and multi-GPU systems.

The overview presentation by Elster and the paper on Throughput Computing on Future GPUs (PDF for personal use) by Hovland and Elster, the on-line PDF files of the presentations from the industrial track together with the four papers above a should provide a great resource for those who want to take a closer look at how one can harness the great computational power of modern GPUs.