Computational Science Community Wiki

Differences between revisions 4 and 5
Revision 4 as of 2010-11-25 14:13:42
Size: 1337
Editor: MichaelBane
Comment:
Revision 5 as of 2010-11-25 15:22:52
Size: 1495
Editor: RobinPinning
Comment:
Deletions are marked like this. Additions are marked like this.
Line 3: Line 3:
Graphical Processing Units (GPUs) used to be for processing graphics for computer screens. Much of these processing was able to be carried out concurrently on small amounts of data. GPUs thus became capable of carrying out the same operations on different data at the same time - a known parallel processing pattern (the SIMD model). The jump was made to using the compute capability of GPUs to carry out not just processing of graphical information but of the main computation itself. This is what is currently know as GPGPU (general purpose GPU) programming, more commonly known as just GPU programming. Graphical Processing Units (GPUs) used to be for processing graphics for computer screens. Much of these processing was able to be carried out concurrently on small amounts of data. GPUs thus became capable of carrying out the same operations on different data at the same time - a known parallel processing pattern (the SIMD model). The jump was made to using the compute capability of GPUs to carry out not just processing of graphical information but of the main computation itself. This is what is currently know as GPGPU (general purpose GPU) programming, more commonly known now as just GPU programming.
Line 5: Line 5:
== What's Available to Buy? ==
Various manufacturers have different hardware to sell. Here's just a brief round-up of the some more modern choices
== What's Available? ==
Various vendors have product lines suitable for general purpose computing. Here's just a brief summary. Feel free to add and amend.
Line 8: Line 9:
Suggestion: always use the product number (eg C2050) to clearly identify. Check the "compute capability" (you'll want 2.0 or higher)
  i. fermi architecture: 512 stream processors, optional ECC
  i. product streams:
   a. tesla - high-end workstation card. Tesla 20-series includes: C2050/M2050/S2050 & C2070/M2070/S2070 (card/rack-mounted?/server respectively)
   a. quadro - workstation/desktops card
   a. !GeForce - desktop card
  * Fermi is an architecture designation (a follow on from the G80 and then GT200, sporting 128 and 240 stream processors): 512 stream processors, optional ECC, 2:1 DP support ([[http://www.realworldtech.com/page.cfm?ArticleID=RWT093009110932&p=7|vs. 8:1 for GT200]])
  * Relevant NVIDIA product lines:
   * Tesla - high-end workstation card. Tesla 20-series includes: C2050/M2050/S2050 & C2070/M2070/S2070 (Workstation '''C'''ard/Node-e'''M'''bedded/breakout '''S'''erver respectively)
   a. Quadro - High end workstation graphics card
   a. !GeForce - Consumer graphics card
Line 15: Line 15:
  i. Radeon
  i.
  * Radeon
  *  

GPUs: What are they?


Graphical Processing Units (GPUs) used to be for processing graphics for computer screens. Much of these processing was able to be carried out concurrently on small amounts of data. GPUs thus became capable of carrying out the same operations on different data at the same time - a known parallel processing pattern (the SIMD model). The jump was made to using the compute capability of GPUs to carry out not just processing of graphical information but of the main computation itself. This is what is currently know as GPGPU (general purpose GPU) programming, more commonly known now as just GPU programming.

What's Available?

Various vendors have product lines suitable for general purpose computing. Here's just a brief summary. Feel free to add and amend.

  • NVIDIA
    • Fermi is an architecture designation (a follow on from the G80 and then GT200, sporting 128 and 240 stream processors): 512 stream processors, optional ECC, 2:1 DP support (vs. 8:1 for GT200)

    • Relevant NVIDIA product lines:
      • Tesla - high-end workstation card. Tesla 20-series includes: C2050/M2050/S2050 & C2070/M2070/S2070 (Workstation Card/Node-eMbedded/breakout Server respectively)

      • Quadro - High end workstation graphics card
      • GeForce - Consumer graphics card

  • AMD/ATi
    • Radeon
  • Intel
    1. Larrabee - potential for the future
  • Others