Sign in with
Sign up | Sign in

Meet The FirePro V9800

AMD FirePro V9800 4 GB: Eyefinity Meets Professional Graphics
By

We went into depth on the FirePro V8800 and Quadro 5000 back in September with Workstation Shootout: Nvidia Quadro 5000 Vs. ATI FirePro V8800.

For the most part, AMD’s FirePro V9800 is very similar to the V8800, with a handful of notable exceptions. Both cards employ AMD’s Cypress GPU with all 1600 stream processors enabled. The V8800’s core clock is set to 825 MHz, while the V9800 set to 850 MHz (that 25 MHz difference should be negligible in most benchmarks).

AMD's "High-End" Lineup
FirePro V9800
FirePro V8800
FirePro V7800
GPU
Cypress
Cypress
Cypress
Stream Processors
1600
1600
1440
Single-Precision TFLOPS
2.72
2.64
2.02
Memory
4 GB GDDR5
2 GB GDDR5
2 GB GDDR5
Memory Bandwidth
147.2 GB/s
147.2 GB/s
128 GB/s
I/O
6 x DisplayPort
3-pin Stereo
Frame Lock/Genlock
4 x DisplayPort
3-pin Stereo
Frame Lock/Genlock
1 x DL-DVI
2 x DisplayPort
Frame Lock/Genlock
Form Factor
Dual-Slot
Dual-Slot
Single-Slot
Power
199 W
208 W
138 W
MSRP
$3499
$1499
$799


They also both sport GDDR5-based frame buffers running at the same 1150 MHz (yielding the same 147.2 GB/s of memory bandwidth on the same 256-bit bus). Granted, the V8800 includes 2 GB of memory, while the V9800 comes equipped with 4 GB. But don’t let twice the memory fool you—it’s actually pretty difficult to take advantage of that much RAM unless you’re working with scientific/oil/gas data sets or 4K frames in stereoscopic 3D.

The two cards occupy dual-slot form factors and support CrossFire. But while the FirePro V8800 only needs two six-pin power connectors, the V9800 requires one six-pin and one eight-pin connector. Moreover, the V8800 offers a quartet of full-sized DisplayPort outputs and a three-pin stereo connector. The V9800 fully utilizes Eyefinity through six mini-DisplayPort connectors (the same stereo plug is available, too).

Eyefinity

Because the FirePro V9800 uses DisplayPort exclusively, there are bound to be questions about hooking up to DVI-based monitors. The good news is that AMD includes five DisplayPort-to-single-link DVI adapters and one DisplayPort-to-dual-link DVI adapter with the card. The single-link dongles support resolutions of up to 1920x1200, while the dual-link dongle enables 2560x1600 on one display.

Of course, configuration is a lot simpler if you go with native DisplayPort monitors, like the Dell P2210Hs on my test bench. Connecting six of them is a simply matter of running mini-DisplayPort-to-DisplayPort cables between the components.

Perhaps best of all (especially for professional users), most of the DisplayPort-oriented bugs we encountered back in March with AMD’s Radeon HD 5870 Eyefinity 6 Edition card have seemingly been worked out. No more monitors disappearing from our Display Group. No more corrupt cursors persisting on-screen. And no more driver-reported failures. The only issue we encountered was a tendency for our 30” Dell 3007WFP to not receive a signal during boot-up (attached to a dual-link active adapter), forcing a manual reboot (Update: We confirmed with AMD that this problem is isolated to the Dell 3007WFP in our lab. Newer models from Dell shouldn't have this issue).

Display all 41 comments.
This thread is closed for comments
Top Comments
  • 20 Hide
    eugenester , November 17, 2010 3:27 AM
    Nvidia shines in the industrial/commercial/scientific market, with their driver team and CUDA/GPGPU tech. Too bad the V9800 fell short of expectations. Also, Nividia cards are obviously going to have better results in Adobe Mercury since both companies worked together on hardware optimization. AMD needs to be more aggressive in working together with software makers, (including games!) to have a stronger hold on both the CPU and GPU markets. Overall, a good read.
Other Comments
  • 20 Hide
    eugenester , November 17, 2010 3:27 AM
    Nvidia shines in the industrial/commercial/scientific market, with their driver team and CUDA/GPGPU tech. Too bad the V9800 fell short of expectations. Also, Nividia cards are obviously going to have better results in Adobe Mercury since both companies worked together on hardware optimization. AMD needs to be more aggressive in working together with software makers, (including games!) to have a stronger hold on both the CPU and GPU markets. Overall, a good read.
  • 7 Hide
    reprotected , November 17, 2010 3:36 AM
    It would actually make sense if they compared with the V8800 and the Quadro 6000. We also need a review of the Quadro 4000, 2000 and the 800, along with the lower Firepro 3D series.
  • 1 Hide
    radiovan , November 17, 2010 3:43 AM
    reprotectedIt would actually make sense if they compared with the V8800 and the Quadro 6000. We also need a review of the Quadro 4000, 2000 and the 600, along with the lower Firepro 3D series.

    1. Definetly, a review of the "lower end" cards would be nice.
    2. Plus, it would be nice to see how well the SLi cards scale.
    3. Also, with the updated (e)nVidia desktop cards (GF100 to GF110), will the Quadro ones see a revision too - if so, when?
  • 1 Hide
    Anonymous , November 17, 2010 4:17 AM
    Benchmark with gpu base render engin like mental images IRay or Chaos Group V-Ray RT
  • 7 Hide
    cangelini , November 17, 2010 4:28 AM
    radiovan1. Definetly, a review of the "lower end" cards would be nice.2. Plus, it would be nice to see how well the SLi cards scale.3. Also, with the updated (e)nVidia desktop cards (GF100 to GF110), will the Quadro ones see a revision too - if so, when?


    Good question (3), I'll ask!

    Cheers,
    Chris
  • 6 Hide
    Cwize1 , November 17, 2010 5:18 AM
    This was a rather underwhelming test suit. I think the fundamental problem you have is that most of the tests you ran were CPU based.

    What most of these production apps use the GPU for is on the fly rendering. For example, sculpting in blender can tax the GPU quite nicely given enough vectors. Another good blender one would be playing back a super resolution baked fluid simulation in real time. For example, take the tom's hardware logo you had before, turn it into water and let the water fall onto a flat surface. Bake the simulation with a ridiculous resolution (as much as you can before blender crashes) and then play the simulation back in real time while watching 5 high definition videos at the same time.
  • -1 Hide
    tony singh , November 17, 2010 5:46 AM
    What a disappointment, hopefully next firepro will be aa winner.
  • 0 Hide
    hell_storm2004 , November 17, 2010 7:06 AM
    The FirePro has still the long way to go to catch up with the Quadro. I hope ATI makes good progress in the workstation models soon like they have come a long way in the desktop market.
  • 0 Hide
    wa1 , November 17, 2010 8:55 AM
    cool...
  • -1 Hide
    eclecticfortune , November 17, 2010 10:41 AM
    In this article's conclusion appeared this statement:
    "If you’re a creative professional working with Adobe’s CS5 suite, then the Quadro is hands-down a no-brainer."
    Benchmarks indicate that the lower priced GTX 480 is a far better choice (cost effective)for those taking advantage of the Mercury Playback Engine running Premiere Pro CS5.
    http://ppbm5.com/Benchmark5.html
    At the top of this page, click on the "MPE Performance Chart" to get a comparison between the different Nvidia Cards with Premiere Pro CS5.
  • -5 Hide
    kristoffe , November 17, 2010 11:11 AM
    I use a 2gb palit 460 and it is amazingly fast. the main problem is the heat, so I took an old heatsink that was short, dremeled it into 9 pieces (3x3) and then used xtreme tuner HD to keep the fan at 70-90% while cuda or dx11 gaming. the hacks for cuda in premiere and other programs work without faking it into a quadro.

    now at intense 1920x1200 gaming i reach 65-72*C instead of average gaming up to 85*C.

    to keep the heatsinks on i used some gap filler 2 part paste from berquist. worked wonders to transfer heat and keep it on.

    get a 3 sli mobo and boom you're off to the races with 2 of these nicely spaced apart. $3500 is overkill unless you need that 5-6th monitor. 4 is great for me.
  • -5 Hide
    nebun , November 17, 2010 11:13 AM
    i will stick with me water cooled 480sli set up for a while, lol
  • 0 Hide
    ohseus , November 17, 2010 12:43 PM
    Is the show in the demo trailer a real one? if not it should be, looks cool.
  • 1 Hide
    eaclou , November 17, 2010 12:52 PM
    I agree that you really need to find a benchmark that measures viewport performance in common 3D DCC applications, and sculpting programs like zBrush or Mudbox.

    There's absolutely no need to include rendering in mentalray, which does not use a GPU at all. At this point in time, there are few GPU renderers that are widely used or as flexible as the established CPU renderers. In the future the GPU will likely make a big difference in rendering times, but for now, most professionals are much more interested in the speed improvements one can get while working in the viewport (i.e. how many polygons / textures can the card display on screen and how quickly.)

    Maybe set up a turntable of an ultra high poly scene (once with and once without textures) and measure the framerate?
  • 1 Hide
    kelemvor4 , November 17, 2010 1:15 PM
    Why are you comparing the top end AMD workstation card to the middrange nvidia Quadro 5000. You did the same thing last time you wrote a workstation review. Nothing like throwing some major slant into the review by not comparing apples to apples (e.g. Quadro 6000). Was it a deliberate slant or just a major error?
  • -1 Hide
    saint19 , November 17, 2010 1:32 PM
    Working station is another thing. nVidia makes the rules and the industries follow that rules, Quadro is the best GPU for workstations.

    Now, I have the same question: Why that performance aren't in Fermi?
  • -1 Hide
    rohitbaran , November 17, 2010 2:22 PM
    Look at the shrunk picture in the article link. It makes the card seem so puny!
  • 1 Hide
    falchard , November 17, 2010 3:02 PM
    The benchmark suite is not very good for comparing Professional cards. Nearly every benchmark was CPU dependant. No shocker in Adobe, AMD gets trounced showing AMD's lack of nVidia's Proprietary GPGPU language.

    Obviously Viewperf is the only real benchmark in the suite and that does not show any trends that are surprising. AMD excels at Maya.
  • 0 Hide
    dragonsqrrl , November 17, 2010 3:55 PM
    greghomeSo..........When are we gonna get Cayman Firepro benchmarks?

    lol... the consumer gaming derivatives haven't even come out yet. The pro cards usually come out within six months of the gaming cards, so I think we'll probably see the Firepro versions of Cayman in the first half of 2011.
  • 1 Hide
    dragonsqrrl , November 17, 2010 4:25 PM
    kelemvor4Why are you comparing the top end AMD workstation card to the middrange nvidia Quadro 5000. You did the same thing last time you wrote a workstation review. Nothing like throwing some major slant into the review by not comparing apples to apples (e.g. Quadro 6000). Was it a deliberate slant or just a major error?

    Perhaps they didn't have a Quadro 6000 on hand? They do admit to the price discrepancy in the conclusion, so it's not like they're trying to cover it up or deceive the viewer into thinking these cards are in the the same price range. I think it's interesting that the Quadro 5000, which can easily be found for $1700 by the way, performs better then the $3500 Firepro V9800 in the majority of benchmarks.

    However it would be very interesting and informative to include the Quadro 6000 in some benchmarks, as it's closer to the V9800 price range and supposedly performs notably better then the 5000.
Display more comments