Jump to content
HWBOT Community Forums

Shrekogre

Members
  • Posts

    77
  • Joined

  • Last visited

Posts posted by Shrekogre

  1. No, I don't have problem with bclk. Running in IDE mode with my SSD I've tested up to 130.

     

    Lower or higher divider (than 1169) for the IGPU results in lower fps when I try to match e.g. 1400MHz with the 1169 option. However with 1169 option and 3:32 mem divider max benchable is 1415Mhz for the GPU, because max stable mem frequency for this APU is 1290. Although I can load windows at say 1330 it is not benchable - crashes on desktop or reboots when I try to run 3dmark. Nothing helps - voltage, timings... I used ~1.7V (more was not needed) for the RAM, because it is the IMC failing. I gained a little on LN2, but not as much as I hoped.

     

    I see top guys run the same multi, so I guess it's a common thing and I need a chip with better IMC.

     

    On a side note - the board is rock stable. Bios recovers very well from bad settings and had absolutely no problem during the first sesion. Great job Asrock and Nick!

     

    Also as Nick said - get a proper cooling for the pwm. Even after several hours at -190 for the cpu and -70 on the memory, the pwm heatsink was still warm and I had a fan placed over it.

     

    BTW, thanks for the bios tip, that saved my @ss. I gues that block at the beginning is just a code for the bios flashing program. Flashed fine with Flashrom on Asus M4A89GTD.

     

    my score gets higher when i start low until 1169. after that it is downhill all the way.

     

    guess my ssd doesn't like the high bclk.

     

    and yeah i was pulling my hair on the bios corrupt. i was darn happy when i could save mine. glad u can save yours too.

  2. I have problems at 15001430+ GPU even with the tool from Stilt. FPS drops no matter what.

    Running Catalyst 13.9 on Win 7 x64.

     

    PS: I get normal scores only at "1169" multi, but my APU's IMC is bad, so I'm limited to 120 bclk :/

     

    same deal with what i have. there seems to be a black hole from 121 to 127 bclk for me. 128 boots but not stable.

     

    Running ide mode helps in bclk. although my ssd drop off after 105. i had to use a pcie 1x 2 port sata card for my ssd to run higher bclk. or running normal hdd helps too.

  3. I have tried that in Sunday without success.

    Board stops at F2 and there's some activity on the USB drive (formatted in FAT32 and the only file in the root directory is the bios).

    After 1-2 minutes it moves to 7C and usb activity led stops, but the board does not boot.

     

    Do you have another amd board? I tried to save my bios with a Asrock A85 ITX with success. not sure if you can use another brand.

     

    Should be able to use the dos or window afuxxx. the only thing you need to do is to trim the bios file from 8196kb to 8192kb, if not it will always fail flashing due to rom size not matching with file size. need to delete the first 0-1000h of the file.

     

    --- Open the bios file with your hex editor. Highlight everything from offset 0000000 to 00001000 and delete it, moving the existing code up. The existing code should just read as FF FF FF, etc.

     

    Always hang when flashing the boot blocks though. so i just flash the rest without the boot blocks.

  4. 1. Result is HWBOT valid. But not FM valid cause you are cheating the Tesselation of the benchmark

    2. same as above

    3. System info is not new enough to verify if you are using the windows 8 RTC bug / cheat. If you are running any other OS than win8 this result is still fine. Otherwise update your system info to 4.20. on futuremark site

    4. You are using a beta or new graphics diver, its not FM valid but HWBOT supports the use of any driver

     

    so result with item 3 is fine for submission on window 7?

  5. For whoever might be interested in this. So far, the numbers for MOA look quite good. Even though the amount of seats in the grand final reduced significantly from 32 (16x2) to 15, there were slightly more participants this year. We went from 72 to 79. The amount of score submissions almost doubled as the counter is now at 420-ish for the entire MOA campaign.

     

    I hope these numbers will provide "upper management" with enough evidence that MOA-type competitions can be organised without spending a rediculous amount of money and still create the same amount of exposure and participance.

     

    Just fyi.

     

     

    I prefer live though. leveled the playing field.

×
×
  • Create New...