Department of Computing

3rd UK GPU Computing Conference

The UK GPU Computing Conference is an informal day of talks aiming to span the whole landscape of accelerated computing and heterogeneous manycore computing. Topics of interest include high-performance computing, embedded and mobile systems, computational science, finance, computer vision and beyond. The goal of the event is to build the UK community of expertise, both industrial and academic.


    Wednesday, 14 December 2011


    Goodenough College, Mecklenburgh Square, London WC1N 2AB, Tel 020 7837 8888
    Location on map

Conference programme:

09:30   Coffee and pastries

10:00   Welcome from the organisers

10:05   Keynote talk:
             "Software is key. Intel MIC: GPU level performance with CPU level programmability"
             Jim Cownie (Intel)

11:00   Porting a Fortran Oceanographic code to GPUs; the GNEMO Project
             Andrew Porter, Stephen Pickles and Mike Ashworth (Daresbury Labs)

11:30   Efficient sparse matrix-vector products on Fermi GPUs
             Istvan Reguly and Mike Giles (Oxford University)

12:00   Anisotropic adaptive mesh coarsening and refinement on GPU architecture
             Matthew Potter (JP Morgan)

12:30   Lunch

13:30   The landscape of GPU programming: a view from ARM
             Anton Lokhmotov (ARM)

14:00   A heterogeneous image reconstruction system for clinical Magnetic Resonance
             Grzegorz Kowalik, Jennifer Steeden, Bejal Pandya,
             David Atkinson, Andrew Taylor and Vivek Muthurangu (UCL)

14:30   BarraCUDA - a Fast Sequence Mapping Software using Graphics Processing Units
             Brian Lam, Petr Klus, Simon Lam, Graham Pullan and Giles Yeo (Cambridge University)

15:00   Application acceleration and optimisation with directives on hybrid supercomputers
             Alistair Hart (Cray)

15:30   Break

16:00   Exploiting OpenCL for heterogeneous computing: a case study
             Simon Mcintosh-Smith, Richard Sessions, James Price,
             Amaurys Ibarra, Tsuyoshi Hamada and Felipe Cruz (Bristol University)

16:30   Automatic Parallelisation: Code Generation for GPUs
             Luke Cartey (Oxford University)

17:00   Using Graphics Processors for Real-Time Global Illumination
             Graham Hazel (Geomerics)

17:30   Close

Steering Committee:

    Prof. Mike Giles, Mathematical Institute, University of Oxford
    Prof. Paul H J Kelly, Department of Computing, Imperial College London
    Dr. Graham Pullan, Department of Engineering, University of Cambridge
    Simon McIntosh-Smith, Department of Computer Science, University of Bristol

Local Organisers:

    Dr. Carlo Bertolli, Department of Computing, Imperial College London
    Adam Betts, Department of Computing, Imperial College London
    Nathan Chong, Department of Computing, Imperial College London
    Prof. Paul H J Kelly, Department of Computing, Imperial College London
    Georgios Rokos, Department of Computing, Imperial College London

Previous Events:

    1st UK CUDA Developers' Conference, Oxford, 2009
    2nd UK GPU Computing Conference, Cambridge, 2010

Contact us:




Last update 16.12.2011