Results 1 to 21 of 21

Thread: Finally 173 Fsb on Soltek, but 3Dmark problem.

  1. #1
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22

    Finally 173 Fsb on Soltek, but 3Dmark problem.

    Yeah Soltek Drv5 with 1.5 Bios finally were kind enuff to let me O/C it to 173 i can do further, and i will soon.
    Sandra shows 2605, so i'm in 2.6 k club
    But i made it work only thru CHIP skew and setting it to 650 ps. I dunno about that much, just my friend said try that. It worked sandra said u r at 173 by bandwidth score, ain't it cool, to be stable And my video card Oc'd to 305/635 TI 4400. So before at 142 FSB i had my 10900 score on 3dmark 2001. With this 173x10 2.8 v mem 1.8 vcore my and CHIP skew at 650 i get only.........drum thing goin'.........6500 ( jaws drop ). I don't get HOW f...ing come it happens... but it's not goin' over 6510.
    only 4 first games or all TEST it just says 6500.
    SO i have this problem here. Also where i have my CHIP skew in BIOS there's
    PCI skew
    AGP skew
    and
    CPU skew.

    Thanks in advance for ANY help.

    any suggestions gentlemen ?

  2. #2
    Hardass
    Guest
    That is strange, hang in there and we will try to get a answer for you.

  3. #3
    Xtreme Legend
    Join Date
    Jun 2002
    Location
    Belgium / Denmark / Russia
    Posts
    2,254
    hmmm could you show us your 3D project compare link or post the framerates ? i'd like to see where you're losing all those points.... if you go back to your old settings and 140 FSB, does the score go up?

    I've never had any of those boards and i swear I have no clue what that chip skew thing does. I was never a fan of soltek boards though. i'm also out of AMD benching a bit lately.... i'm sure DDTUNG , OPP or Icee could be of more help to you ... but for now, i'd say show us the exact detailed framerates.

    Also, try running some other benches.. PCmark, super pi, as long as its CPU intensive that'd be fine;..

    good luck

  4. #4
    Xtreme Enthusiast
    Join Date
    Sep 2002
    Location
    UK
    Posts
    875
    I think I had something similar with CPU Decode (can't remember exactly what it's called atm - options Ultra/Fast/Normal) set to Ultra at high fsb, was getting around 4000 marks lower than expected. Try setting it to Fast.

  5. #5
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    ok i had 142 before and yeah without any CHIP skew. it worked ok, 11 k with my oc'd TI4400. I can show u 142 FSB benches without skew only monday. But my 173 one's i'll post as fast as i can. The reason why I did CHIP skew is because without CHIP skew of 550 ps it won't POST at 166 mhz fsb or more. PC works stable sandra's benches work perfectly, 2600 on mem bandwidth, and 4830 on arithmetic. i'll go try PCmark, then.

    Thanks a lot for replyin'.

  6. #6
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    CPU Score 5180
    Memory Score 3579
    HDD Score N/A

    CPU Tests


    Jpeg Decoding 17.1 MPixels/s
    Zlib Compression 6.8 MB/s
    Zlib Decompression 80.7 MB/s
    Text Search 177.3 KRounds/s
    Audio Conversion 96.7 KB/s
    3D Vector Calculation 70.8 FPS

    Memory Tests


    Raw Block Read - 3072KB 1477.0 MB/s
    Raw Block Read - 536KB 1478.8 MB/s
    Raw Block Read - 384KB 1508.2 MB/s
    Raw Block Read - 48KB 20721.6 MB/s
    Raw Block Read - 6KB 20315.2 MB/s
    Raw Block Write - 3072KB 811.2 MB/s
    Raw Block Write - 1536KB 811.3 MB/s
    Raw Block Write - 384KB 835.0 MB/s
    Raw Block Write - 48KB 13027.5 MB/s
    Raw Block Write - 6KB 12825.2 MB/s
    Raw Block Modify - 3072KB 694.3 MB/s
    Raw Block Modify - 536KB 680.2 MB/s
    Raw Block Modify - 384KB 657.1 MB/s
    Raw Block Modify - 48KB 6958.6 MB/s
    Raw Block Modify - 6KB 6981.1 MB/s
    Random Access - 24576 (1536KB) items 926.2 MB/s
    Random Access - 12288 (768KB) items 934.4 MB/s
    Random Access - 6144 (384KB) items 939.0 MB/s
    Random Access - 1536 (96KB) items 4428.8 MB/s
    Random Access - 768 (48KB) items 5346.4 MB/s
    Random Access - 96 (6KB) items 6114.6 MB/s
    Video - 1 line 64.1 FPS
    Video - 4 lines 63.6 FPS
    Video - 16 lines 65.2 FPS
    Video - 32 lines 64.0 FPS

    Hard Disk Tests


    Cached File Write N/A
    Uncached File Write N/A
    Cached File Read N/A
    Uncached File Read N/A
    File Copy N/A

    Crunch Tests


    Jpeg Decoding 9.1 MPixels/s
    Raw Block Modify - 3072KB 274.9 MB/s
    Blit Test 52.6 FPS

    Pro Version - Windows XP 2D Tests


    Lines 1183.1 lines/s
    Ellipses 710.0 ellipses/s
    Arcs 1068.4 arcs/s
    Beziers 817.7 beziers/s
    Curves 203.3 curves/s
    Filled Ellipses 3865.5 ellipses/s
    Gradient Boxes 144.5 rectangles/s
    Gradient Polygons 23.8 polygons/s
    Filled Pies 2581.9 pies/s
    Images 170.8 images/s
    Cached Bitmaps 586.7 cached bitmaps/s
    Strings 217125.0 characters/s


    a little later 3dmark scores.

  7. #7
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    CPU AMD Athlon(tm) XP/MP/4 1744 MHz 1900+
    FSB 173 MHz
    Memory 512 MB

    Graphics Chipset NVIDIA GeForce4 Ti 4400 305/635
    Driver Name NVIDIA GeForce4 Ti 4400
    Driver Version 6.13.10.3082
    Video Memory 128 MB

    Program Version 3DMark2001 SE
    Resolution 1024x768 32bit
    Texture Format Compressed
    FSAA Disabled
    Z-Buffer Depth 24bit
    Frame Buffer Double
    Rendering Pipeline D3D Pure Hardware T&L





    3DMark Score 6494 3D marks

    Game 1 Car Chase - Low Detail 116.6 FPS
    Game 1 Car Chase - High Detail 53.9 FPS
    Game 2 Dragothic - Low Detail 85.4 FPS
    Game 2 Dragothic - High Detail 45.1 FPS
    Game 3 Lobby - Low Detail 98.5 FPS
    Game 3 Lobby - High Detail 51.8 FPS
    Game 4 Nature 23.7 FPS

    Fill Rate (Single-Texturing) 617.6 MTexels/s
    Fill Rate (Multi-Texturing) 618.2 MTexels/s

    High Polygon Count (1 light) 27.9 MTriangels/s
    High Polygon Count (8 lights) 12.5 MTriangels/s

    Environment Bump Mapping 86.6 FPS
    DOT3 Bump Mapping 67.2 FPS

    Vertex Shader 15.2 FPS
    Pixel Shader 57.5 FPS
    Advanced Pixel Shader 37.3 FPS

    Point Sprite 32.5 MSprites/s

  8. #8
    Xtreme Mentor
    Join Date
    Jun 2002
    Location
    PA, USA
    Posts
    2,662
    wow man, I can't speak for the PCMark score, but that 3DMark is terrible for ur setup....hope someone thinks up the prollem, as I'm at a loss....
    E8600. EP45-UD3P. 500FSB 24/7. Still WinXP. Heatware

  9. #9
    Xtreme Enthusiast
    Join Date
    Sep 2002
    Location
    UK
    Posts
    875
    Since I've got nothing better to do at the moment than sit here watching Cliffy Cam for the UT2K3 demo release , I've had a quick look at the manual for that mobo.

    The setting I was referring to is called System Performance in your bios, in the DRAM Clock/Drive Control menu. If this is set to Fastest (= Ultra on my system) change it to Fast.

  10. #10
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    it is on fast, and uhh what do u think of this suggestion


    Maybe you put your videocard back to normal speeds and test it then... it could be that it can't the memory speed and automatically lowers other memory timings.

  11. #11
    Xtreme Enthusiast
    Join Date
    Sep 2002
    Location
    UK
    Posts
    875
    That's a good idea. Going too far with the gfx o/c can also lower the score, but when it's happened to me it's only been by a few hundred points, not thousands. Try it anyway.

  12. #12
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    i don't think so cause now i have only 6400 :-)
    hmmm
    maybe i should play with skews orr hmmm interesting.

    and lowering the fsb does same sh*t.

    problem is in video card i think.
    Last edited by Lionsgraphics; 09-13-2002 at 08:04 PM.

  13. #13
    Xtreme Enthusiast
    Join Date
    Jun 2002
    Location
    Bulgaria
    Posts
    706
    Originally posted by Lionsgraphics

    Fill Rate (Single-Texturing) 617.6 MTexels/s
    Fill Rate (Multi-Texturing) 618.2 MTexels/s

    Those are really poor Fill rate scores.... And I mean REALLY POOR!
    Mollov,
    Runnin' a 700MHz SlotA Thunderbird...

    HardwareSight.com Owner

  14. #14
    Xtreme Enthusiast
    Join Date
    Sep 2002
    Location
    UK
    Posts
    875
    Did you try lowering the gfx card overclock or not? If that doesn't improve things, try disabling AGP Fast Writes and AGP Master 1 WS Read/Write in the bios, these make minimal difference anyway.

  15. #15
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    that's what i said the score goin' DOWN when i lower my gfx clock.
    i still think it's skew problem...
    normally my system won't boot at 166 mhz or more (that's when 1/5 divider kicks in ) it'll boot only 600 ps or more. hmmmmm.
    or smthn happened with my gfx i'll try 142 fsb monday without any skews, we'll see some results, but damn i love that sandra's results they kick ass ;-)

  16. #16
    Xtreme Enthusiast
    Join Date
    Sep 2002
    Location
    UK
    Posts
    875
    Sorry, didn't read it correctly.

    I've just been checking some of my 3dmark projects and TodB is right, those are really terrible Fill Rate scores. My old GF2 got 381/725, my Ti4600 gets 1217/2553!

    Also, in the game tests Car Chase and Lobby seem sort of ok, but Dragothic and Nature are way too low. These are exactly the symptoms of a GF4 that's clocked too high. But you've tried lowering it, so I dunno what it is.

    The only other thing I can think of trying is lowering the System Performance to Normal (from Fast) and see if that makes any difference.

  17. #17
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    ok check this out ;-)
    for those who doesn't believe i did what u said ;-)

    Operating System Microsoft Windows XP
    DirectX Version 8.1

    Mobo Manufacturer
    Mobo Model VT8367-8233
    AGP Rates (Current/Available) 4x / 1x 2x 4x

    CPU AMD Athlon(tm) XP/MP/4 1680 MHz 1900+
    FSB 166 MHz
    Memory 512 MB

    Graphics Chipset NVIDIA GeForce4 Ti 4400 275/550
    Driver Name NVIDIA GeForce4 Ti 4400
    Driver Version 6.13.10.3082
    Video Memory 128 MB

    Program Version 3DMark2001 SE
    Resolution 1024x768 32bit
    Texture Format Compressed
    FSAA Disabled
    Z-Buffer Depth 24bit
    Frame Buffer Double
    Rendering Pipeline D3D Pure Hardware T&L





    3DMark Score 5952 3D marks

    Game 1 Car Chase - Low Detail 105.6 FPS
    Game 1 Car Chase - High Detail 51.6 FPS
    Game 2 Dragothic - Low Detail 78.6 FPS
    Game 2 Dragothic - High Detail 40.1 FPS
    Game 3 Lobby - Low Detail 89.5 FPS
    Game 3 Lobby - High Detail 47.5 FPS
    Game 4 Nature 21.6 FPS

    Fill Rate (Single-Texturing) 559.6 MTexels/s
    Fill Rate (Multi-Texturing) 559.6 MTexels/s

    High Polygon Count (1 light) 26.3 MTriangels/s
    High Polygon Count (8 lights) 11.3 MTriangels/s

    Environment Bump Mapping 74.0 FPS
    DOT3 Bump Mapping 60.5 FPS

    Vertex Shader 14.6 FPS
    Pixel Shader 52.4 FPS
    Advanced Pixel Shader 34.0 FPS

    Point Sprite 28.5 MSprites/s



    OK so again i'm sure it's a skew sh*t, thanks for suggestion Mr. Lavender.
    GRRRR
    Any other suggestions ? ;-)

  18. #18
    Xtreme Enthusiast
    Join Date
    Sep 2002
    Location
    UK
    Posts
    875
    Nope, I'm out of ideas...

  19. #19
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    and i did the normal thing, instead of fast. thanks for effort man.


    damn where r moders ;-) ?

  20. #20
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    and i did the normal thing, instead of fast. thanks for effort man.


    damn where r moders ;-) ?

  21. #21
    Registered User
    Join Date
    Aug 2002
    Location
    Portland, Maine
    Posts
    22
    Much like its predecessor, the SL-75DRV, the SL-75DRV4 includes a series of functions that allow users to insert timing delays into a number of circuits that are critical to Overclocking. The goal of using these delays, though, is not to augment performance, but rather to improve the stability of an Overclocked system. Below, we've included a list of the affectable circuits, and the delays that can be set, as measured in picoseconds



    CPU skew adjust: 50ps to 750ps in steps of 50ps

    CHIP skew adjust: 50ps to 750ps in steps of 50ps

    PCI skew adjust: 100ps to 1500ps in steps of 100ps

    AGP skew adjust: 100ps to 300ps in steps of 100ps


    hmmmm....

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •