Announcement

Collapse
No announcement yet.

poor g400max 3dmark2000 score

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

  • poor g400max 3dmark2000 score

    my system
    celly 400@619mhz
    64mb ram pc100 cas 3
    g400max oc'174/217 agp 2x103mhz bus 206mhz
    my score is 2070 in 1024x768 16b 16z tb
    309/308 mpixel fillrate
    108mpixel env. bumpmap

    my problem is cpu or g400 settings??????
    please post you resoult or reg setting

  • #2
    I think your results aren't too bad. I'm pretty sure 3DMark 2000 uses the PIII SSE instructions, which might lower a Celeron score a bit. Also, 3DMark 2000 gives much lower scores than 3DMark 99 anyway.
    On a PIII o/c to 560 and G400 o/c to 155 I get 2643 3DMarks (with the default test settings), 277 for fill-rate, and 96 for environmental bump-mapping which I think is about on par for a system like mine.

    Comment


    • #3
      For 1 get 64mb more. Totally ness for any hardcore gamer, as is obvious by your stats.
      Also the sse use in 3dmax 2000 gives an unfair number to us celery users. when in real life the results should be closer

      ------------------
      Damnit Jim I'm a film maker, not a systems tech!
      Damnit Jim I'm a film maker, not a systems tech!

      Comment


      • #4
        Hi Kenobi,

        I disagree with you. If SSE is used properly (like it is in 3dmark, TurboGL, and DX7), then there is a HUGE difference. Just ask Kruzin what a difference SSE makes.

        Rags

        Comment


        • #5
          I disagree to you RAGS. In todays games there is no or small performance boost with SSE. I have celery 400 OC to 450. I tried P3 500 for a while and I was major dissapointed. Only real incerase was in sintetic benchmarks like 3D mark. Talking about TGL? Check this out: celery 450 52 fps, P3 500 61 fps. (Q3, 640*480 billinear, excatly same settings). In my favorite resolution (800*600*32) result was THE SAME for both processors: 44 FPS. So what happened? I returned P3, and I am waiting for new SSE celery 600.

          Comment


          • #6
            Hi

            Just adding some benchmarks

            3DMark2000 score: 3518
            Q3 1024/768, Bilinear, otherwise default: 36.1
            G400Max OC'ed to 110%

            And to sasa:
            Was it a plain P3-500 or a coppermine?
            Because I just don't see a major difference between a Celery SSE (128K On-Die)and PIIIE (256K On-Die) except for the pricing!

            ------------------
            Abit BH6, P3-600E@800(6x133), 128mb MCT 8ns, G400MAX, SBLIVE, AHA 2940AU, WDC 4.3, WDC 6.4, IBM GXP34 20.5, Mitsumi FX40, Yamaha 4416S, Zyxel Prestige 100IH, 3Com 905B-TX On W98SE, DX 7.0a, PD 5.52

            Abit BE6-2 (Rev 2.), P3-1000E@1050(10x105/3), 768mb Kingston 7.5ns CAS3, G400MAX, SBLIVE, AHA 2940AU, IBM GXP75 60Gb (*2), IBM IBM GXP75 45Gb, Mitsumi FX48, Yamaha 4416S, Zyxel Prestige 200, 3Com 905C-TX On W98SE Lite, DX 8.1, PD 6.82

            Comment


            • #7
              It was an old P3, with 512 Kb cache.
              I agree with you, and I expect low prices of new celery while they supposed to have similar performance like coppermine.
              Your 3D mark score is quite good. Last month I installed GeForce DDR and my score was 3650 (celery 450). I wonder how high would be GeForce score with that fast procesor?

              Comment


              • #8
                TTcharon, what are your settings for 3DMark? I get about 2600 with a P3-550E o/c to 770 and a G400Max o/c 110% (256Mo) RAM. This in 1024x768, 32bpp, 32bzb.
                In 3DMark 2000 at these resolution the tests are fill-rate limited, except (maybe) the high quality tests, which only count for 1/3 of the result...


                ------------------
                Corwin the Brute

                Corwin the Brute

                Comment


                • #9

                  To sasa:

                  I have seen the GeForce (DDR version) hit just below 5000 marks and the normal version tops about 4000 marks with a P3-800E.

                  To Corwin_Brute:

                  My test were run in 1024x768, 16bpp, 16bzb.
                  Or in other words, the unregistered 3Dmark2000 default benchmark setting.

                  What about your Q3 fps?


                  ------------------
                  Abit BH6, P3-600E@800(6x133), 128mb MCT 8ns, G400MAX, SBLIVE, AHA 2940AU, WDC 4.3, WDC 6.4, IBM GXP34 20.5, Mitsumi FX40, Yamaha 4416S, Zyxel Prestige 100IH, 3Com 905B-TX On W98SE, DX 7.0a, PD 5.52

                  Abit BE6-2 (Rev 2.), P3-1000E@1050(10x105/3), 768mb Kingston 7.5ns CAS3, G400MAX, SBLIVE, AHA 2940AU, IBM GXP75 60Gb (*2), IBM IBM GXP75 45Gb, Mitsumi FX48, Yamaha 4416S, Zyxel Prestige 200, 3Com 905C-TX On W98SE Lite, DX 8.1, PD 6.82

                  Comment


                  • #10
                    sasa what Rags was saying is that WHEN IMPLEMENTED PROPERLY, SSE makes a big difference. Most games have little or no optimal SSE optimization, hence the small difference in scores
                    [size=1]D3/\/7YCR4CK3R
                    Ryzen: Asrock B450M Pro4, Ryzen 5 2600, 16GB G-Skill Ripjaws V Series DDR4 PC4-25600 RAM, 1TB Seagate SATA HD, 256GB myDigital PCIEx4 M.2 SSD, Samsung LI24T350FHNXZA 24" HDMI LED monitor, Klipsch Promedia 4.2 400, Win11
                    Home: M1 Mac Mini 8GB 256GB
                    Surgery: HP Stream 200-010 Mini Desktop,Intel Celeron 2957U Processor, 6 GB RAM, ADATA 128 GB SSD, Win 10 home ver 22H2
                    Frontdesk: Beelink T4 8GB

                    Comment


                    • #11
                      I just ran 3DMark 2000 again, with the default settings, here are the results :
                      Final score : 3411 (SSE enabled)
                      Helicopter test :
                      - low : 69.5
                      - medium : 48.1
                      - high : 21.1

                      Adventure test :
                      - low : 65.1
                      - medium : 49.1
                      - high : 31.4

                      Quake 3 (demo 1), TGL installed :
                      - fastest : 99.6
                      - fast : 88.2
                      - normal : 75.6
                      - high : 43

                      My system :
                      P3 550E@770
                      256Mo CAS 3
                      G400Max@110% (AGP 1x for stability)
                      everything SCSI

                      So, SSE DOES make a difference...

                      ------------------
                      Corwin the Brute

                      Corwin the Brute

                      Comment


                      • #12
                        To Corwin the Brute:

                        Yes it Does! In 3DMark2000 Anyway!

                        I'm running AGP 2x and don't have any stability problems, what mb are you using?

                        By the way, Pretty OK 3DMark Benchmark you have!

                        In Q3, what do mean by:
                        fastest?
                        fast?
                        normal?
                        high?

                        TTCharon

                        ------------------
                        Abit BH6, P3-600E@800(6x133), 128mb MCT 8ns, G400MAX, SBLIVE, AHA 2940AU, WDC 4.3, WDC 6.4, IBM GXP34 20.5, Mitsumi FX40, Yamaha 4416S, Zyxel Prestige 100IH, 3Com 905B-TX On W98SE, DX 7.0a, PD 5.52

                        Abit BE6-2 (Rev 2.), P3-1000E@1050(10x105/3), 768mb Kingston 7.5ns CAS3, G400MAX, SBLIVE, AHA 2940AU, IBM GXP75 60Gb (*2), IBM IBM GXP75 45Gb, Mitsumi FX48, Yamaha 4416S, Zyxel Prestige 200, 3Com 905C-TX On W98SE Lite, DX 8.1, PD 6.82

                        Comment


                        • #13
                          Heh. That wacky benchmark gives me a paltry 1400 marks at 'default benchmark' setting. This on a Celeron 533, 128MBs, G400Max (nothing o/ced)which has no difficulty throwing Q3A around my screen at maximum detail, 1024x768x32.

                          The funniest is the so-called 'CPU-Test', whose frame-rate monitor reads a constant 7.50 fps, regardless of the fact that one can plainly see that this is at best wildly inaccurate.

                          Crikey! I hadn't realised what a pile of old shite my machine must be! I'd better tell all my other stuff to stop running so well - according to MadOnion, it's impossible :-)

                          Unless there's some kind of hidden 'Make System Look Good under 3dMark' switch in the registry which no-one's told me about...:-)

                          Comment


                          • #14
                            Sasa: You did the wrong upgrade. I had a celery 300@450 and when I went K7 *500* my jaw keeped droping.

                            About q3 at that res, it is fill-rate limited. It will do ~44 fps in a celery 450 or with a coppermine 1000.

                            Comment


                            • #15
                              My motherboard is an Asus P3B-F. I have stability issues when running 3D apps with a 140MHz FSB... Runs fine up to 124, problems at 133 and beyond.
                              Actually, my processor can run with a 150MHz FSB, but half of my RAM can't handle that kind of speed.

                              About SSE, Q3Arena is also SSE enabled. For Q3, fastest, fast, normal and high are the different default settings (resolution, texture quality...). I play with my custom settings, but I prefer to use Id's defaults for benchmarking purpose.


                              ------------------
                              Corwin the Brute

                              Corwin the Brute

                              Comment

                              Working...
                              X