Announcement

Collapse
No announcement yet.

G400 +gaming becoming disappointing

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • G400 +gaming becoming disappointing

    Let me post my entire system specs before everyone jumps the gun on this post

    My System
    Athlon 650
    MSI 6167 motherboard(no superbypass)
    256MB PC100
    Maxtor 30GB 7200RPM
    Matrox G400MAX
    Sound Blaster Live! X-Gamer
    D-Link 10/100 nic
    USRobotics 56k faxmodem (ISA)
    Generic 32x CDRom, Floppy
    Antec MidTower 300watt case
    Camabridge FPS2000 Speakers
    Logitech Mouseman Wheel+
    Some controllers

    Right now I'm running WindowsME build 2525
    with the Matrox 5.52.015 and TurboGL 1.3(believe me, it runs better than 2000 or 98se)
    Heres some timedemo scores (Quake2 and Quake3 built-in timedemos)
    -----------------
    Quake III Fastest - 71.3 fps
    Quake III Normal - 64.6 fps
    Quake III HQ - 39.6
    Each test was run 3 times. The BEST of the 3 scores were taken
    -----------------
    Quake 2 1024x768x16 - 71.5 fps
    x32 - 65.9 fps
    Quake 2 800x600x16 - 107.6 fps
    x32 - 102.4 fps
    Didn't bother running lower resolutions.

    I would like to note that running Quake3 at 1024x768 in 32bit resulted in a frame rate of 29.6 fps. Running in 16 bit at the same resolution resulted in a 46.5 frame rate. The visual quality wasn't even that great, I have to play with the geometric detail one notch down and the texture quality one notch down unless I want an even slower game.

    I'm not expecting this card to compete with a GeForce. Just a TNT2Ultra. And from what I can see, it pales in comparison.

    Dues Ex won't even run anywhere near 30fps, MDK2 averages about 40-45 (1024x768 16bit) but takes huge drops in open areas.

    Is there something that I don't know about this card? Its starting to annoy me to the point where selling it looks real good and for the money I could get selling it a Viper II would be more than enough to take its spot.

    One more gripe about the card. I had heard numerous reports of the G400 being a good overclocker, so I decided to try. It wouldn't go more than 5% without noticable visal artifacts on the desktop and in games. Don't know why, but it was very sensitive. Most cards aren't that touchy.

    If anyone has any tweaking secrets or options to get this thing running at least a little faster, plz email me or post a message back on the board.

    Thanks

    Velocity

  • #2
    If you have trilinear filtering enabled, then you really aren't do a fair comparison between the G400 and the TNT2, the TNT2 fudges the trilinear using a dithering type process, which is not much diferent than bilinear. If you want to compare the two, run the TNT2 with trilinear on, and the G400 with bilinear, you will get identical visuals, and a fair comparison.

    Rags



    ------------------
    Partnership for an idiot free America

    Comment


    • #3
      Only test using trilinear filtering was the High Quality test in Q3A. All others used bilinear.

      Velocity

      Comment


      • #4
        And your bitching about 60+ fps in Q3A? Get real.



        ------------------
        Partnership for an idiot free America

        Comment


        • #5
          Your´re doing something wrong there. Here are some numbers in win2k:
          Q2 timedemo 1 1024x768x16: 78.6 fps
          Q3 1.17 HQ settings demo001: 46.5 fps

          Athlon 700, Gigabyte 7IX (No superbypass also), 128 Mb SDRAM, G400 32 Mb DH oc´ed to 150/200 (MAX settings) using 5.10 beta win2k drivers, SB Live 1024 player using liveware 2 for win2k.

          Do you have 751 AGP miniport drivers installed?

          PS: Above benchmarks using 16 bit Z-buffer (32 bit Z gives artfacts in Q3. As fas as I can remember same settings win 32 bit Z-buffer in win98 gave something like 44 fps)


          [This message has been edited by Nuno (edited 27 June 2000).]

          Comment


          • #6
            Hey rags,

            Haven't you learned by now that we just have to have that 2-3fps for our games to look good? But then again, you will always have those few that spend more time playing benchmarks than the actual games themselves.

            Joel
            Libertarian is still the way to go if we truly want a real change.

            www.lp.org

            ******************************

            System Specs: AMD XP2000+ @1.68GHz(12.5x133), ASUS A7V133-C, 512MB PC133, Matrox Parhelia 128MB, SB Live! 5.1.
            OS: Windows XP Pro.
            Monitor: Cornerstone c1025 @ 1280x960 @85Hz.

            Comment


            • #7
              Nuno,

              I like you u=you chose the HQ setting. Nobody uses those. Change the res to 1024X768 with 32bc and 32bt, Trilinear,
              geometric and everything to High. Let's see those 21fps scores in NT5!
              C:\DOS
              C:\DOS\RUN
              \RUN\DOS\RUN

              Comment


              • #8
                DosFreak: I used the HQ setting just to compare with Velocity numbers. Reading his post I assume he is using Q3 default settings (Fastest, Normal, HQ): I think HQ 39 fps is low for an Athlon 600/Max.

                Personally I play 1024x768x32, 16 bit tex, bilinear, everything else to the max. It benches 34.6 fps. Playable.

                Everything to the MAX gives 22.7 fps. Yes, maybe 21 fps with a 32-bit z-buffer (wich is not playable due to texture corruption - marks in the walls)

                Comment


                • #9
                  Woohoo! Someone finnally confirmed 22fps in NT5! Geesh..for months now I've been trying to see if anyone was getting the same score. Thanks!
                  C:\DOS
                  C:\DOS\RUN
                  \RUN\DOS\RUN

                  Comment


                  • #10
                    Velocity,

                    I was told by Matrox not to use the 5.52 PD drivers with the VIA chipset. They recomended using the 5.41 or 6.00beta drivers. Also, after loading the latest VIA 4.22 4in1 drivers and AGP 4.03 patch everthing worked fine. Hope this helps.....

                    Comment

                    Working...
                    X