Computational Science Community Wiki

GPU Club Meeting: 9 December 2011

alt="NVIDIA CUDA Research Centre"

  • Software for GPUs inc. compiler/directives, maths libs & tools (debuggers and profilers)

About 50 people attending the following presentations

After the presentations the following points were discussed and actions agreed.

On the Available Resources:

  1. The ACM has online GPU courses (link TBC)

On the presentations:

  1. Questions arose about GPU-izing other sections of the ParaFEM code. It was something the speaker was investigating. Moving the entire solver to the GPU was the aim.
  2. Could one GPU per CPU core be used in multicore/multi-gpu systems. Possibly but usually more cores than GPUs and currently still limited by the PCI bottleneck.
  3. On the use of directives, were they considered good enough in terms of performance seen for the amount of time spent adding them to the code (rather than the perhaps more time-consuming process of writing CUDA/OpenCL kernels from scratch). The generally were considered good enough and something the application scientists would be more likely to learn.

  4. The use of FPGAs vs GPUs presentation led to some discussion about using other low-power devices to drive the GPUs (rather than use a CPU such as x86). Texas Instruments and Arm devices were suggested.
  5. It was asked whether other groups in the University were using FPGAs. One group is looking in to capturing data from the power network (link TBC).

On community involvement:

  1. Michael reported he'd had offers from CAPS to present about their HMPP directives and Cray to present about their own work in this area. They'll be asked along to separate meetings of the GPU Club next semester.

  2. From the floor, attendees requested speakers covering the following technologies:
    • OpenCL on Arm architecture
    • AMD Bulldozer
    • MIC from Intel
    • High-level languages that combine MPI, GPU etc
  3. It was requested that anyone attending GPU conferences report back to the group. Michael to report on the 3rd UK GPU Computing Conference he was attending at Imperial on 14/12/11.

  4. Maksims Abalenkovs kindly offered to help with the organisation of the next GPU Club meeting

GPU Club Meetings

Previous Meetings

  • Tues 26 Nov 2013: 2-3pm, B8 George Begg. Christian Obrecht on GPU implementations of fluid dynamics simulations on regular meshes: some recent advances

  • Weds 13 Nov: 2pm, Univ Place. John Michalakes (NOAA) and Craig Davies (Maxeler Dataflow)

  • Weds 30 Oct: Intermediate CUDA training run by NVIDIA

  • Tues 29 Oct: 2pm, Univ Place, NVIDIA and Stephen Longshaw.

  • Weds 2 Oct 2013 - Large Scale Optimization and High Performance Computing for Asset Management, Daniel Egloff (QuantAlea)

  • Tuesday 23 July MathWorks (GPUs for MATLAB) and NVIDIA (GPUs & CUDA)

  • Thur 2 May 2013 Lessons from GTC and on using the Intel Xeon Phi

  • Mon 10 Dec 2012 Dataflow and MultiGPU SPH

  • Tues 25 Sept Seminar on implementing financial models on GPUs, FPGAs and in the Cloud

  • Mon 15 Oct: OpenCL training from UoM IT Services

  • Thurs 25 Oct: Hands-on "OpenACC" workshop run by Cray UK Ltd.

  • 17 May 2012 Speakers on healthcare policy simulation in OpenCL, MHD algorithms in CUDA, Tridiagonal Solvers in CUDA

  • 20 April 2012 Francois Bodin, CAPS: "Programming Heterogeneous Many-Cores using Directives" using HMPP

  • 23 March 2012 Roko Grubisic, ARM: "Embedded Computer Graphics and ARM Mali GPUs"

  • 02 March 2012 Speakers on profiling, sparse matrix algebra and atmospheric chemistry

  • 09 Dec 2011 MPI and GPUs, directives-based programming, FPGA and GPU comparison, ideas for 2012

  • 30 Sept 2011 GPU programming in FORTRAN, multiple GPUs, image reconstruction

  • 15 July 2011 Jack Dongarra key note on Emerging Technologies

  • 18 Mar 2011 OpenCL, debugging and profiling tools, porting C to CUDA, real time analysis

  • 26 Nov 2010 biological MD, smoothed particle hydrodynamics, Monte Carlo financial models, Markov models