Computational Science Community Wiki

Differences between revisions 1 and 153 (spanning 152 versions)
Revision 1 as of 2010-09-22 10:55:39
Size: 1054
Editor: MichaelBane
Comment:
Revision 153 as of 2016-10-03 10:59:03
Size: 4276
Editor: LouiseLever
Comment:
Deletions are marked like this. Additions are marked like this.
Line 1: Line 1:
= GPU Club = #acl LouiseLever:read,write,admin,delete,revert RcsStaffGroup:read,write All:read

= University GPU Club =
Line 3: Line 5:
''Initial thoughts'' <<Include(GPU/contents)>>
----
Line 5: Line 8:
== why, what, who ==
Many people on campus now starting to use GPUs and a few have used extensively. This Club could meet quarterly for people to present and exchange ideas and mainly to make contacts. Back it up by comm wiki page for discussion & tips.
The GPU Club is an initiative to bring people together to share ideas and experiences in order to more effectively overcome any barriers to getting their research done. Since November 2010, the Club has set up a [[#EmailList|email list]] with 300 members, held over [[#PreviousMeetings|15 meetings]], ran a [[GPU/meetings/23July2013|GPU Day]] with collaboration talks with key vendors, organized 2 sets of training, set up a [[#SIG|SpecialInterestGroup]], summarised [[GPU/UniversityResearchers |GPU research at UoM]] and has successfully negotiated with vendors regarding training and hardware provision.
Line 8: Line 10:
Could promote OpenCL & multicore programming tips/techs too == Introduction ==
Graphics Processing Units ([[../|GPUs]]) are now established as an option for undertaking intensive computation tasks without having to buy expensive compute nodes with fast interconnects. Recently, other "many core" technologies, particularly the [[http://wiki.rac.manchester.ac.uk/community/Technical/hardware/XeonPhi|Xeon Phi]] from Intel and a resurgence in interest in FPGAs, have emerged.
Line 10: Line 13:
People:
 * Ben R - experienced & just purchased Fermi
 * James, George, Jonathan, MKB
 * Ser Huan Poon - has somebody who's used them before and they hold research mtgs
 * Neil Burton - Perri doing PhD to port code to GPU
 * Paul Popelier - EPSRC HPC s/w dev bid to port DL-MULTI/POLY to GPU
 * Dongarra/Silvester/Milan - EPSRC HPC s/w dev bid to develop linear algebra packages for GPUs
 * Mike Croucher - experience of MATLAB on GPUs
The GPU Club brings together The University of Manchester staff and students to exchange best practice and real world examples on emerging technology.
Line 19: Line 15:
Kit
 * DISCO (tesla)
 * RQ viz (?)
 * Ben (fermi)
 
##If one is to believe the sales pitches from the main vendors, NVIDIA and AMD/ATi, it’s possible to have your own personal supercomputer with 448 cores capable of peak performances of a teraflop for just £3,578 - about half the total performance of the University’s old “Horace” service. (These are 2010 prices online ex-VAT price for Amari 7046GT-TRF with a single NVIDIA C2050 card. Single precision performance. Performance has increased greatly and prices have dropped recently)
Line 25: Line 17:
== launch afternoon ==
speakers, networking & agreement to do again
And to address questions such as:
 * Is this peak performance achievable for real life codes?
 * What programming methods are best supported by these cards?
 * How much effort is involved in porting?
 * NVIDIA's GPU or OpenCL? NVIDIA or AMD/ATi? GPU or [[http://wiki.rac.manchester.ac.uk/community/Technical/hardware/XeonPhi|XPhi]]? Or does it matter?
 * Where can I get my hands on one to see if it is for me?
Line 28: Line 24:
potential speakers
 * Mike Croucher
 * Ben R
 * SHP's person
 * DISCO folk (eg on GPU porting of DL-POLY)
There is a lot of interest across all faculties on campus concerning GPUs. Many people are seriously considering their use for their future research. Several groups have [[GPU/UniversityResearchers|already]], or are in the process of, writing and running models on GPUs.

<<Anchor(EmailList)>>
== Join the Mailing List ==
''For various reasons it has recently been decided that the university GPU Club Mailing List should be for UoM staff and students only. Any non-UoM email addresses have been removed from the Mailing List and will be used for occasional announcements.''<<BR>><<BR>>
We have a mailing list for UoM staff & students to share news and ask questions relating to GPUs. To join you need to log in to listserv and search for GPUCLUB (or [[http://listserv.manchester.ac.uk/cgi-bin/wa?REPORT=&n=gpuclub&m=&a=&s=0&X=1AD8630577A063B0CC&z=3|click here]]), select the GPUCLUB email list then select Subscribe from the drop down list beneath and click on 'Submit Changes'. Full instructions are given at the bottom of the page.<<BR>><<BR>>
For non-UoM email addresses, please contact <<MailTo(michael DOT bane AT manchester DOT ac DOT uk)>>.

<<Anchor(mtg)>>
##[[attachment:GPU/Club/Meetings/23Mar2012/GPUClub_23Mar2012_poster.png|{{attachment:GPU/Club/Meetings/23Mar2012/GPUClub_23Mar2012_poster_small.png|poster|align="right"}}|&do=get]]

<<Anchor(PreviousMeetings)>>
<<Include(GPU/Club/Meetings/info)>>

## ''Please click and download poster (right) for display''

Please [[HelpOnLogin|login]] and add your suggested topics and speakers for future meetings:
 * Directives-based programming for GPUs
 * Is GPU HPC?
 * GPU-enabled computational libraries
 * Comparison of GPU technologies for research codes
 * MD/many-body interactions, including ACEMD and SPH
 * comparison of OpenCL and NVIDIA's GPU implements
or <<MailTo(michael.bane@manchester.ac.uk,email us)>> with your suggestions


<<Anchor(SIG)>>
== Special Interest Groups ==
 * [[/RealTime | Real Time Applications Using GPUs]]

University GPU Club


alt="NVIDIA CUDA Research Centre"

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


The GPU Club is an initiative to bring people together to share ideas and experiences in order to more effectively overcome any barriers to getting their research done. Since November 2010, the Club has set up a email list with 300 members, held over 15 meetings, ran a GPU Day with collaboration talks with key vendors, organized 2 sets of training, set up a SpecialInterestGroup, summarised GPU research at UoM and has successfully negotiated with vendors regarding training and hardware provision.

Introduction

Graphics Processing Units (GPUs) are now established as an option for undertaking intensive computation tasks without having to buy expensive compute nodes with fast interconnects. Recently, other "many core" technologies, particularly the Xeon Phi from Intel and a resurgence in interest in FPGAs, have emerged.

The GPU Club brings together The University of Manchester staff and students to exchange best practice and real world examples on emerging technology.

And to address questions such as:

  • Is this peak performance achievable for real life codes?
  • What programming methods are best supported by these cards?
  • How much effort is involved in porting?
  • NVIDIA's GPU or OpenCL? NVIDIA or AMD/ATi? GPU or XPhi? Or does it matter?

  • Where can I get my hands on one to see if it is for me?

There is a lot of interest across all faculties on campus concerning GPUs. Many people are seriously considering their use for their future research. Several groups have already, or are in the process of, writing and running models on GPUs.

Join the Mailing List

For various reasons it has recently been decided that the university GPU Club Mailing List should be for UoM staff and students only. Any non-UoM email addresses have been removed from the Mailing List and will be used for occasional announcements.

We have a mailing list for UoM staff & students to share news and ask questions relating to GPUs. To join you need to log in to listserv and search for GPUCLUB (or click here), select the GPUCLUB email list then select Subscribe from the drop down list beneath and click on 'Submit Changes'. Full instructions are given at the bottom of the page.

For non-UoM email addresses, please contact <michael DOT bane AT manchester DOT ac DOT uk>.

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

Please login and add your suggested topics and speakers for future meetings:

  • Directives-based programming for GPUs
  • Is GPU HPC?
  • GPU-enabled computational libraries
  • Comparison of GPU technologies for research codes
  • MD/many-body interactions, including ACEMD and SPH
  • comparison of OpenCL and NVIDIA's GPU implements

or email us <michael.bane@manchester.ac.uk> with your suggestions

Special Interest Groups