1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Nvidia GPU for PCI X16 1.0

Discussion in 'General Hardware' started by GJS1, 11 May 2010.

  1. GJS1

    Gangster

    Joined: 13 Aug 2009

    Posts: 316

    Hey, I have a system I want to put a GPU in for 24/7 running of a sicintific application like [email protected] but don't want to have to upgrade the motherboard on it - I'm pretty dure this is a X16 just not 2.0 and was wondering what GPUs I could have supported by this system?
    Thanks (PS I don't want to have to upgrade the mobo)
     
  2. Alex74

    Mobster

    Joined: 7 May 2009

    Posts: 3,541

    Location: London

    budget? Also which motherboard model are we talking here?
     
  3. J.B

    Soldato

    Joined: 16 Aug 2006

    Posts: 5,922

    why do you need a graphics card at all, it's only going to use power up.

    Connect to it using RDP or something like that?

    WHat is the make and model of the motherboard or post a picture.
     
  4. cmndr_andi

    Capodecina

    Joined: 23 Apr 2009

    Posts: 12,040

    Location: Cheshire

    I imagine that the OP is planning to make use of the incredible parallel processing capability of a powerful graphics card to power his distributed computing system. Since a GPU when run properly can be orders of magnitude faster than a CPU in these kinds of tasks, it does make sense. Here is the CUDA based version of [email protected], they advertise it as:

    As for the graphics card running in a PCIe v1 x16 slot - this will be fine. This slot has the same bandwidth as a PCIe v2 x8 slot - and according to these tests the performance dip in a top-end GTX 480 is tiny.

    However, I would suggest against supporting the SETI program, bad things could happen :D
     
  5. Redmint

    Mobster

    Joined: 28 Jan 2010

    Posts: 3,190

    Location: Guildford

    Haha :p but yes CUDA based software can run pretty quick compared to the equivalent CPU software.