1. This forum section is a read-only archive which contains old newsgroup posts. If you wish to post a query, please do so in one of our main forum sections (here). This way you will get a faster, better response from the members on Motherboard Point.

Re: Fastest graphic card for Windows workstation use (2D, not gaming)

Discussion in 'Matrox' started by DaveW, Dec 28, 2007.

  1. DaveW

    DaveW Guest

    You just read the answer to your question on Tom's charts.
     
    DaveW, Dec 28, 2007
    #1
    1. Advertisements

  2. DaveW

    Mr.E Solved! Guest


    What the hell are you going on about? Every time you "unlock" Windows?
    Are you posting via Babelfish?

    If you are using a specialty application that requires a Quadro, you
    should have half a clue more than you do. If you do not, you are wasting
    everyone's time.

    I say spend the $3699 and have the fastest 2d-windows unlocking
    experience this side of DOS.
     
    Mr.E Solved!, Dec 29, 2007
    #2
    1. Advertisements

  3. DaveW

    Paul Guest

    The OPs original posting mentions Vista. Perhaps the confusion is
    over Aero compositing. If the machine was coming out of standby,
    the video card doesn't have power when the computer is sleeping,
    and the video card needs to be reloaded from the ground up. All those
    composited windows would need to be loaded from system memory,
    or even re-rendered. In my mind, that is not a "2D thing". Something
    entirely different.

    *******
    For some "2D fun", try a benchmark like this old timer:

    "WinTune 98 1.0.43"
    http://comunitel.tucows.com/win2k/adnload/37681_30039.html

    Leave just the "Video Test" selected and let it run three times.
    These are my results, on a 9800Pro and a 3.1GHz P4.

    Summary
    RADEON 9800 PRO -
    1280x1024@32bits/pixel
    290±0.42(0.14%) Video MPixels/s

    Video Details

    AccOpt: Normal
    Total video time (s): 3.6
    Window open time (s): 0.0033
    Text scroll time (s): 0.029
    Line drawing time (s): 1.9
    Filled objects time (s): 0.44
    Pattern blit time (s): 0.0032
    Text draw time (s): 0.5
    DIB blit time (s): 0.78
    Window close time (s): 0.017

    Presented more for its comedy value than anything else. There was
    a time when results like that mattered. It'd be interesting to see
    what someone with a powerful system can manage for comparison.

    I tried to find a later version of that benchmark, but haven't managed
    to find a download.

    Paul
     
    Paul, Dec 29, 2007
    #3
  4. DaveW

    Fred Guest

    Here you go
    C2duo E6600 running XP

    Summary
    Radeon X1950 Series
    1280x1024@32bits/pixel
    340±1.4(0.4%) Video MPixels/s

    Video Details

    AccOpt: Normal
    Total video time (s): 3.1
    Window open time (s): 0.005
    Text scroll time (s): 0.18
    Line drawing time (s): 1.5
    Filled objects time (s): 0.28
    Pattern blit time (s): 0.0012
    Text draw time (s): 0.8
    DIB blit time (s): 0.36
    Window close time (s): 0.0037
     
    Fred, Dec 29, 2007
    #4
  5. * Luca Villa:
    Yeah, right. Manufacturers websites as the reference. Now *thats* a
    reliable source....NOT
    Funny, the site that opens on my webbrowser doesn't talk about 2D
    performance but multimonitor setups: "We got our hands on a several
    multimonitor graphics adapters and threw them at a mishmash of monitors
    of different sizes and resolutions to see if our personal video wall
    could really improve our productivity"
    Where does the article say that?
    Ever thought why no-one is talking about 2D performance any more nor why
    2D performance hasn't been benchmarked by reputable magazines and
    hardware sites for ages? Again for you: 2D performance of the last ~8
    years or so is *more* than fast enough for *anything* 2D, period. That's
    a fact. And if you understand how 2D acceleration works i.e. under
    Windows and why the bandwidth needed for 2D is incredible low, much
    lower than even the cheapest crap gfx card provides, then you know why
    no-one talks about 2D performance any more.

    BTW: things like video decoding support (MPEG2/HDTV etc) is *not* part
    of the 2D performance. In fact, video hardware support has basically
    *nothing* to do with gfx performance. It's done by a separate part of
    hardware that is integrated in todays GPUs.
    No, it can't. The waiting time has nothing to do with the gfx card. If
    you logon to Windows the appropriate user profile has to be loaded.
    Especially if you're on a network (ADS) this can take several seconds
    because the local Windows has to retrive user data from the server. Even
    on a standalone PC this can take some time, depending on disk
    performance, CPU and memory. The gfx card simply does shit about that.

    You came here for an advice and you got it. If you don't believe us fine
    then go ahead and buy the most expensive gfx card that you can find if
    you think you will getter 2D performance. But I'd recommend you get at
    least a basic understanding how these things really work.

    Benjamin
     
    Benjamin Gawert, Dec 29, 2007
    #5
  6. Then skip the expensive gfx card (That won't help here) and get more ram and
    a faster CPU (That WILL help).
     
    Thomas Andersson, Dec 29, 2007
    #6
  7. DaveW

    Augustus Guest

    You came here for an advice and you got it. If you don't believe us fine
    You're wasting your time with this guy. He obviously didn't come here for
    real advice and is 100% equipped with unbendable preconcieved opinions not
    based on fact. Even worse, not willing to read and learn when shown actual
    facts. Let him go out and get his Quaro and FireGL.
     
    Augustus, Dec 29, 2007
    #7
  8. * Luca Villa:
    I think you probably refer to the PC Mark 05 results. Well, it might
    look to you that way but just reading some numbers without having a clue
    what has been tested doesn't help.

    FYI:

    - the "2D Graphics Memory" tests are testing the bandwidth of the gfx
    memory for 2D copy operations. The relevance to real world applications
    is *zero* because memory bandwidth isn't a limiting factor for 2D for
    almost a decade now.

    - The "2D WMV video playback" test has a mis-leading title: it doesn't
    test 2D performance but the performance of video playback, done by
    simply playing back a HDTV video (1920x1080) with Windows Media Player
    at maximum possible frame rate. It doesn't tell you *anything* about 2D
    application performance, it just tells you how well HDTV videos can be
    played back by Windows Media Player. This test is just nonsense as
    todays cards often support HDTV hardware playback with certain players
    or additional software, so basically this test is useless.

    - "2D Transparent Windows" creates 30 Windows with a sweeping "fading"
    effect (alpha blending). The number just tells you how many of these
    Windows can be created per second, it's not only affected by the gfx
    hardware but also by the driver and even by what other processes are
    running on the computer. While this test at least has some remote
    relevance to real work (Window drawing) it also has no real world
    relevance as you never ever see or notice the difference between a
    system that can draw 3800 of these windows per second or "just" 2800.

    Mind you, understanding hardware is no idiot's game where you just have
    to compare some numbers. If you don't know what exactly has been tested,
    how this stuff works and interacts and what also influences the results
    you can't read anything from the numbers.

    Benjamin
     
    Benjamin Gawert, Dec 29, 2007
    #8
  9. * Luca Villa:
    if you want ressources then do your homework. But you better dig very
    deep because you won't find any somewhat recent reliable source for a
    topic that is a no-brainer for almost a decade now.

    You came here with your question and this question has been answered by
    people that very obviously know much more than you do about hardware in
    general. If you don't believe in what we tell you why did you even came
    here to ask? You probably are way better when buying the most expensive
    gfx card you can find. At least it saves you the trouble about using
    your brain.

    Augustus, I think you're right. We're just wasting our time.

    Benjamin
     
    Benjamin Gawert, Dec 29, 2007
    #9
  10. DaveW

    Riffrafter Guest

    Bingo!

    Faster processing under Vista for standard "Desktop" Windows apps will be
    *much* more impacted by a faster processer and plenty of RAM (2GB minimum).
    Couple that with a decent $100 video card and you'll be in very good shape.

    I have a middle of the road Dell with an AMD X2 5200+ CPU and 4GB RAM. From
    "locked" desktop to everything back and ready to work is under 2 seconds.
    From Sleep mode to everything back (desktop and all open windows apps ready
    to go) is under 5 seconds.

    I also have an NVIDIA 8800GT, but that only helps me in 3D gaming (and boy
    does it help!), but I got essentially the exact same desktop response using
    the 7300LE that originally came with my system. The 7300 LE probably costs
    less than $50 today and it handled Aero and standard windows apps under
    Vista very well...

    -Riff
     
    Riffrafter, Dec 29, 2007
    #10
  11. DaveW

    Mr.E Solved! Guest

    You need to find another hobby, ciao!
     
    Mr.E Solved!, Dec 29, 2007
    #11
  12. DaveW

    JLC Guest

    I'm really starting to think this guy is a troll. Just seems that no matter
    how the facts are presented to him, he has some silly response. And poor Ben
    has spent a hell of a lot of time trying to help this guy, and all he gets
    back is more BS. JLC
     
    JLC, Dec 30, 2007
    #12
  13. DaveW

    Dima Guest

    Here you go, download and run the BitBit 2D benchmark.
    http://www.karpfenteich.net/colorful/bitblt.html

    couldn't resist, but hope it's accurate
    HD 3870 results - 1680x1050, Vista x64

    BitBlt:
    avg: 2461.0 fps [2884.0 MB/sec]
    max: 3975.1 fps [4658.3 MB/sec]
    min: 97.6 fps [114.4 MB/sec]

    ReverseBlt:
    avg: 727.9 fps [853.0 MB/sec]
    max: 4292.0 fps [5029.7 MB/sec]
    min: 103.5 fps [121.3 MB/sec]
     
    Dima, Dec 30, 2007
    #13
  14. DaveW

    Dima Guest

    Q6600@2.66Ghz hd3870, P35 P5K Dlx
    1680x1050
    AccOpt: Normal
    Total video time (s): 4.9
    Window open time (s): 0.019
    Text scroll time (s): 0.93
    Line drawing time (s): 0.32
    Filled objects time (s): 0.28
    Pattern blit time (s): 0.58
    Text draw time (s): 1.8
    DIB blit time (s): 0.94
    Window close time (s): 0.0047

    1024x768
    AccOpt: Normal
    Total video time (s): 2.3
    Window open time (s): 0.014
    Text scroll time (s): 0.37
    Line drawing time (s): 0.14
    Filled objects time (s): 0.064
    Pattern blit time (s): 0.17
    Text draw time (s): 1.2
    DIB blit time (s): 0.35
    Window close time (s): 0.0042
     
    Dima, Dec 30, 2007
    #14
  15. * Luca Villa:
    Yes. The only difference is that these professional 2D cards (Quadro
    NVS/FireMV) are certified for certain professional 2D applications and
    that these cards unlike consumer cards (Geforce/Radeon) support big
    multihead installations (quad head and more).

    They don't offer a better performance.

    Benjamin
     
    Benjamin Gawert, Dec 30, 2007
    #15
  16. * JLC:
    Well, I was also thinking about that someone who really is interested in
    reality might one day search for this topic with groups.google.com, so
    probably a few facts don't hurt. Of course my also my patience is
    limited and starts to get overstressed.

    Benjamin
     
    Benjamin Gawert, Dec 30, 2007
    #16
  17. DaveW

    Paul Guest

    I find the text results rather curious. Maybe it is due to
    ClearType or something ? My OS is Win2K, and maybe that makes
    a difference.

    http://en.wikipedia.org/wiki/Cleartype

    Paul
     
    Paul, Dec 30, 2007
    #17
  18. * Paul:
    The solution to this riddle is to see relevance (or better: the lack of)
    of BitBlt for 2D performance.

    Benjamin
     
    Benjamin Gawert, Dec 30, 2007
    #18
  19. On a Q6600/2.4GHz/8800GTX - 1680x1050, WinXP Pro 32bits:

    BitBlt:
    avg: 1336.6 fps [1566.3 MB/sec]
    max: 2519.0 fps [2952.0 MB/sec]
    min: 804.9 fps [943.3 MB/sec]

    ReverseBlt:
    avg: 1189.5 fps [1393.9 MB/sec]
    max: 1419.9 fps [1663.9 MB/sec]
    min: 498.5 fps [584.1 MB/sec]

    Weird numbers. Is this is in any way meaningful? :)

    Regards, Patrick.
     
    Patrick Vervoorn, Dec 30, 2007
    #19
  20. DaveW

    Outback Jon Guest

    I would think that numbers that are so ridiculously high above what
    *any* monitor is capable of actually displaying on the screen really
    wouldn't mean much.

    Realistically, if your monitor can only show 120 fps (120 Hz) then
    having the capability to show 10x that on average, and 7x that at a
    minimum probably doesn't really do anything for you.

    As has been stated in this thread before, and will be again, I'm sure,
    the 2D acceleration of graphics cards really has little to do at this
    point with the biggest complaint of users. It's usually processor
    power, lack of memory, or (most likely) Windows crappy coding that is
    responsible for slowdowns on the Windows desktop.

    --
    "Outback" Jon - KC2BNE

    AMD Opteron 146 (@2.8) and 6.1 GHz of other AMD power...
    http://folding.stanford.edu - got folding? Team 53560

    2006 ZG1000A Concours "Blueline" COG# 7385 CDA# 0157
     
    Outback Jon, Dec 30, 2007
    #20
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.