Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums DaVinci Resolve Quadro 4000 slower than 8800 GT?

  • Quadro 4000 slower than 8800 GT?

    Posted by John Sellars on December 10, 2010 at 8:25 pm

    In case any of you are ready to pull the trigger on the new card, see my xbench results below. I got mine at Videoguys for $895. Amazon has it for $834, but out of stock. Don’t have Resolve yet, just getting all the pieces first. I have an email into tech support to see what’s going on. This is just a pure apples to apples test in Slot 1 with driver v256.01.00f03 (latest driver 16 NOV 10).

    QUADRO 4000
    Results 189.09
    System Info
    Xbench Version 1.3
    System Version 10.6.5 (10H574)
    Physical RAM 6144 MB
    Model MacPro3,1
    Drive Type ST31000528AS
    Quartz Graphics Test 227.97
    Line 196.54 13.08 Klines/sec [50% alpha]
    Rectangle 258.42 77.15 Krects/sec [50% alpha]
    Circle 220.73 17.99 Kcircles/sec [50% alpha]
    Bezier 234.67 5.92 Kbeziers/sec [50% alpha]
    Text 239.05 14.95 Kchars/sec
    OpenGL Graphics Test 161.54
    Spinning Squares 161.54 204.93 frames/sec

    8800 GT
    Results 191.33
    System Info
    Xbench Version 1.3
    System Version 10.6.5 (10H574)
    Physical RAM 6144 MB
    Model MacPro3,1
    Drive Type ST31000528AS
    Quartz Graphics Test 231.59
    Line 197.07 13.12 Klines/sec [50% alpha]
    Rectangle 265.27 79.20 Krects/sec [50% alpha]
    Circle 225.66 18.39 Kcircles/sec [50% alpha]
    Bezier 236.41 5.96 Kbeziers/sec [50% alpha]
    Text 244.83 15.32 Kchars/sec
    OpenGL Graphics Test 163.00
    Spinning Squares 163.00 206.77 frames/sec

    Sascha Haber replied 15 years, 4 months ago 3 Members · 3 Replies
  • 3 Replies
  • Sascha Haber

    December 10, 2010 at 9:35 pm

    Well, I got the same results , maybe 10% more with the Gt120 hooked up to the screen and the GTX 285 in Slot 1.
    So the question is, does this say anything about Resolve ?
    I dont trust Xbench…
    It benchmarked my Raid with 190 mb/s
    Resolve with 540 or so…and I can playback two streams of HD…so it seems true.

    Does anyone have a project with a simple sequence or still frame and a bunch of heavy nodes we could benchmark ?

  • Walter Soyka

    December 13, 2010 at 3:10 pm

    I am not blown away by the Quadro 4000’s spec sheet either, but Resolve doesn’t care about OpenGL performance. Resolve uses NVIDIA’s CUDA technology, which allows programmers to use the GPU for number crunching instead of just display.

    This is why the configuration guide specs two graphics cards; one will drive the monitor only, leaving the second card free to handle processing.

    Walter Soyka
    Principal & Designer at Keen Live
    Motion Graphics, Widescreen Events, Presentation Design, and Consulting
    RenderBreak Blog – What I’m thinking when my workstation’s thinking
    Creative Cow Forum Host: Live & Stage Events

  • Sascha Haber

    December 13, 2010 at 3:49 pm

    Yeah, we kinda know that, dont we 😉
    Just kidding…
    The question remains…
    Can anyone with decent DaVinci experience set up a project with a set of nodes that really stress those cores ?
    Then we could start deactivating them and see how our system react.
    Or is there some more sophisticated way to test CUDA and DaVinci performance.
    I mean, dont get me wrong…
    I am more than happy I got the GTX 285 and instead bought the Decklink.
    The Q 4000 looked shiny, but I dont really need it it seems.

    A slice of color…

    DaVinci 7.0.3
    Dual Xeon 2,4
    OSX 10.6.5RAM 6 GB
    RAID 8TB intern
    Extern 6TB
    GTX 285 / GT 120
    WAVE

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy