Heterogeneous Computing

HCSIG
HCSIG - Heterogeneous Computing Special Interest Group

HCSIG is a series of meetings in which the participants discuss recent developments in heterogeneous computing and related research areas. The aim is to hold regular meetings where we can discuss recent publications, technologies, and share knowledge regarding the use of single-node heterogeneous computational resources, (e.g., graphics cards, multi-core CPUs, etc.).

The group consists of participants from, among others, SINTEF ICT, simula.research laboratory, and CMA, who share a common interest in heterogeneous computing. If you want to attend these meetings, send an email to André R. Brodtkorb.

The meetings are held at SINTEF in Oslo (Forskningsveien 1, 0373 Oslo). To attend the meetings, simply send an email to André Rigland Brodtkorb.

Coming Meetings

After a long period of silence, the meetings will now continue. The first meeting will be announced shortly.

Previous Meetings

  • 2009-11-19 -- 10:00-11:30
    Two papers to read this time. The first discusses scalability of operating systems over multiple (heterogeneous) cores. The second paper gives an overview of visualization techniques for scientific data.
    1. Andrew Baumann, Paul Barhamy, Pierre-Evariste Dagandz, Tim Harrisy, Rebecca Isaacsy, Simon Peter, Timothy Roscoe, Adrian Schüpbach, and Akhilesh Singhania, The Multikernel: A new OS architecture for scalable multicore systems
    2. Raphael Bürger and Helwig Hauser, Visualization of Multi-variate Scientific Data
  • 2009-06-02 -- 10:00-11:30
    I have recently read two papers regarding reverse-time migration, which is used to perform seismic imaging. One uses the Cell BE, whilst the other uses the GPU. I found that both contain some interresting points. The two papers are: 
    M. Araya-Polo, F. Rubio, R. de la Cruz, M. Hanzich, J. Cela, and D. Scarpazza, “3D seismic imaging through reverse-time migration on homogeneous and heterogeneous multi-core processors,” Scientific Prog., vol. 17, no. 1-2, pp. 185–198, 2009.
    P. Micikevicius, “3D finite difference computation on GPUs using CUDA,” in Workshop on General Purpose Processing on Graphics Processing Units. New York, NY, USA: ACM, 2009, pp. 79–84.
  • 2009-02-19 -- 10:00-11:30
    We will discuss chapters five ("Exascale Application Characteristics", pages 61 - 82), and eight ("Exascale Challenges and Key Research Areas", pages 207 - 232), from the Exascale Computing Study Report. This sums up to about 45 pages, but the pages are small, and quite readable. The report discusses "the major challenges and the areas where additional targeted research could lay the groundwork for overcoming them" to achieve a 1000X speed increase within 2015.
  • 2009-01-27 -- 10:00-11:30
    1. Nicolaas Groeneboom will give us a short (~30 minutes) technical presentation of Polly (http://www.irio.co.uk/polly), a free C++-based game engine he has developed. See the project webpage, and http://www.vterrain.org/ for more details.
    2. In addition, we will discuss the paper M. Christen, O. Schenk, P. Messmer, E. Neufeld, and H. Burkhart, Accelerating Stencil-Based Computations by Increased Temporal Locality on Modern Multi- and Many-Core Architectures. First International Workshop on New Frontiers in High-performance and Hardware-aware Computing (HipHaC'08)
  • 2009-01-07 -- 10:00-11:30
    We will in this meeting discuss what OpenCL really is, according to the spec. You can find it at http://www.khronos.org/opencl. Now, the spec is a whoopin' 293 pages long, so I suggest we read what we each find most interesting. Some suggestions:
    * The OpenCL architecture
    * The OpenCL C Programming language
    * OpenCL Numerical compliance
    * Image addressing and filtering
    * Optional extensions
  • 2008-11-27 -- 12:00
    1. André will present something on OpenCL from SC08
    2. K. Datta, M. Murphy, V. Volkov, S. Williams, J. Carter, L. Oliker, D. Patterson, J. Shalf, K. Yelick, Stencil Computation Optimization and Autotuning on State-of-the-Art Multicore Architectures, Supercomputing (SC), 2008.
  • 2008-09-18 -- 10:00
    1. L. Seiler et. al., Larrabee: A Many-Core x86 Architecture for Visual Computing, ACM TOG, 2008
    2. Trond and Johan will present some hilights from Intel Developer Forum 08 and NVISION 2008
  • 2008-07-03 -- 10:00
    1. Technical Brief NVIDIA GeForce GTX 200 GPU Architectural Overview
  • 2008-06-10 -- 10:00
    1. A. Eichenberger et al., Optimizing Compiler for the Cell Processor Proc. 14th Int'l Conf. Parallel Architectures and Compilation Techniques, IEEE CS Press, 2005, pp. 161-172.
  • 2008-05-22 -- 10:00
    1. T.Morvan, M. Reimers, and E. Samset, High Performance GPU-based Proximity Queries using Distance Fields, accepted to Computer Graphics Forum.
    2. Sud, Govindaraju, Gayle, Manocha, Interactive 3D Distance Field Computations using Linear Factorization, Proc. ACM Symposium on Interactive 3D Graphics and Games (I3D), 2006.
  • 2008-04-29
    1. D.A. Bader, V. Agarwal, K. Madduri, and S. Kang, High Performance Combinatorial Algorithm Design on the Cell Broadband Engine Processor, Parallel Computing, 33(10-11):720-740, 2007.
    2. Inoue, Hiroshi Moriyama, Takao Komatsu, Hideaki Nakatani, Toshio, AA-Sort: A New Parallel Sorting Algorithm for Multi-Core SIMD Processors, Parallel Architecture and Compilation Techniques, 2007. PACT 2007. 16th International Conference on.

Published December 20, 2010