Jump to content

naddie

Members
  • Content Count

    174
  • Joined

  • Last visited

Posts posted by naddie


  1. Okay - update -

     

    3DMark05 has been running for about an hour now, which is much much MUCH longer than before. In fact, it's still going. Did I solve the issue? If so, what would it be?

     

    My bet is on temperature. Cased up in a Super Lanparty case must be stuffy. And I just found out that the chipset is HOT HOT HOT! Hot to the touch, actually. maybe just changing the factory goop for Artic Silver Ceramiq worked? It was pretty hot before too, but I figured it was within acceptable range if DFI engineers designed it this way.

     

    The only downfall to the layout is the chipset - just not in a good place for ventalation. If it was located by the memory/CPU, it would have recieved extra cooling of my large HSF (huh huh..... large...).

     

    I wonder if I should continue the RMA? Well, I'll leave it overnight running 3DMark and see. But if it fixes it, I'd be faced with a problem - how to get cooling to the chipset.

     

    Oh - also just found out about the Expert under-reporting the CPU temps. I am running at 47*C max according to the nVidia's monitor. So if what they say is true, then I'm probably in the 56*C mark!

     

    However, the entire HSF (even near the base) is just barely warm - even when in the Lanboy case....


  2. Huh..! :drool:

     

    Thats how every mobo is shipped????

    How is that possible? I know NONE of my mobo's were ever shipped with jumper in clear position..!!

    I dont if it will fry the mobo or not(most likely it will), but there is no possible way it can work if you power on with the clear position..!

     

    I know of a PC CHIPS mobo I once bought for a client did. :)

     

    http://www.epinions.com/content_131662188164

     

    That's my epinions of it. It drove me nuts until I found out. If I remember correctly, it just didn't POST. This was the only board that came shipped that way. Unless someone at QA didn't check for it, or someone at NewEgg opened the box, messed with it, and then put it back.

     

    I guess this motherboard started it all - my relentless pursuit of making things work. Well, at least when it comes to computers. :) Anything else, I'd return to the store immediately upon not powering up.


  3. Geez! All this time I thought the Thermalright SI-120 and Silverstone FM121 fan was doing a good job, but it was because of the firmware reporting a lower number!?

     

    Yikes! Scaring me because CPU is reported at 38*C idle, and 47*C under load! It looked right to me all this time! If all this is true (I have no reasons to doubt everyone here, afterall), then I was running my CPU up to 57*C and not know it!

     

    Maybe this is why I was having issues? Then again, primed overnight without issues. Just the 3DMark05 issue I was having in another thread I started.

     

    PS - took the board out to RMA, but while waiting, I set it back up just to test, ina dit seems to be running fine now. Must be the temps?


  4. mine is still crashing in games with the new ballistix. I'm using the default settings for it.

     

    Its strange. the pc gives no indication of anythings wrong then bam, it freezes. I then have to do a hard reset. temps are fine and settings are at default.

    it could be the motherboard.

     

    Oh, for your issue, it may actually be the Ballistix. I read in the Recommended Memory thread that although it was highly recommended, lately they are failing after a couple months of use. You can either RMA them and hope for the best, or just use the G.SKILLs. In the end, I decided on the G.SKILLs HZ.

     

    Of course, now that I am having issues, I am not one to give praise to anything yet. :)

     

    I cross my fingers that the next Expert board will be a saint.


  5. Well the initial impression was very impressive - it was very easy to work with. I figure in the long run, all DFI Lanparty series are like this. But the reason why I like the Expert specificially is the layout of the board.

     

    BTW, here's an interesting thing - after calling and emailing the RMA, the WIndows install went completely through. I installed the chipset drivers from 6.7. I did not use the installer - i just extracted it using WinRAR and installed it individualling via Device Manager. I even installed the video like that (using the latest on nVidia's site).

     

    No nVidia software got installed except for the nVidia's gfx driver (which is so closely integrated with the driver, it IS the driver). No nMixer, no nTune, no nVidia firewall.

     

    Okay - so the only application I installed is 3DMark 05 - and so far, after a couple of restarts, it still doesn't crash. It is looping now. It's going stronger than before. I'll leave it overnight until I get the RMA #.

     

    This is very strange. I'd still RMA it just to be sure.

     

    EDIT: Spoke too soon - it crashed. But it did last much longer than before.

     

    EDIT: Got it out of my case - and now that I did, I took off the factory goop from the chipset and used Artic Silver Ceramiq - and running it outside of the box. LoL. I know by now, most of you would say "You're already RMA, why not quit now?"

     

    I dunno... I guess I just wanna find out if all the lil things I do helped. :) When I get the new board, I will probably do the chipset goop upgrade before I put it in the case.


  6. Well, Robert also emailed me, (rgone), and I just emailed them my info. I guess the RMA process just started. :) Thanks for the quick repsonse, btw, Robert - if you are reading this. :)

     

    On the 6th try, I finally got WinXP installed, and installed all drivers - gonna do a quick 3DMark05 test (and that's all I'm gonna test for my sanity) on stock speeds and voltages before I take this board out and box it up. I am also waiting on the RMA #, so I don't feel like taking it apart until I get it.

     

    I wll update you guys on this. I am NOT going to a different mobo (or at least I dont plan to). I am sticking with the Expert. At least for the forseeable near future.


  7. Yep - Monarch computers wont warrantee it or take it back if you seperated the combo. :( DFI is much better though - there are HUMANS who answers the phone in San Jose. :) Too bad either way, I'll have to take the board out.

     

    I wonder if they want ALL the accessories back. I hope not, cuz that would mean I would also have to undo all of my cabling, which was hard to do in the first place. Try wiring all those drives in a Super Lanboy. -_-;


  8. Okay, I'm thoughrouly pissed and frustrated now.

     

    After trying all the methods, I am now reinstalling Wndows just as a frest start. Guess what? Stock speeds and voltages - WinP just locke dup trying to install itself - in the non-graphical part of it. just stopped on the file "parhmse.inf" at "81%" in the "Setup is copying files..." stage (non graphical).

     

    WTF is going on here? It's begining to feel like a non-video issue. RMA the motherboard? Any of the mods here that may have some insight?

     

    I even thought there may be an issue with 3dmark05 and dual core athlons. But this new development cancels that theory.

     

    What, pray tell, causes instability predictably in 3d05? What causes it to lock up during install of windows? Im gonna try again and start the install - im pretty sure it would go trhough. just weird that it didnt now. But at the same time, incredably resilent to any memtests or priming!!! Runs like champ with those two testing!

     

    Also, I tried 3dmark203 loop and prime x2 (2 instances) - after a while, the whole system hard locks. needs hard restart. Is that suppose to happen (like an incompatibility between 2x prime and 3dmark03?) Can someone check for that to see if 2x prime in different affinity and 3dmark03 can be ran at the sametime without a hard lock?

     

    i appologize for my typing weird and thoughts all over the place. it's late, im tired, and pretty cranky right about now. I've had to use the backspace several times in typing this, lol.

     

    gnite.


  9. I'm currently requiring 1.6v to reach 2.62GHz on my x2 3800+. Too much?

     

    My other question is the chipset voltage. No one asked about that. How much is too much? I had it set at 1.59 as a way to test if my unstable 3DMark05 runs are caused by not enough voltage. I need a guidline as to how far can i take it.


  10. Wel, it seems that I already have the latest eVGA N517 BIOS (5.70). Updated anyways.

     

    I also tried the different versions of drivers. Now using the official one from nVidia's site (non-beta).

     

    I also took the time to replace whatever TIM eVGA used from the factory with Artic Silver Ceramic - same issues.

     

    Reinstalled 3DMark05.

     

    Reinstalled OS.

     

    Ran at stock speeds.

     

    Turned Memory Bank Interleave off and on.

     

    Turned off CPU Throttling. (CnQ was always off though).

     

    Underclocked both the video GPU and memory.

     

     

     

     

     

     

     

     

     

     

    I think I will be RMAing the video card. :(


  11. I was gonna RMA my 7800GT CO, but after tearing it apart and as5'ing the ram chips and the core, it's running just fine and I'm scoring higher and higher as I go. I'd reccomend you do the same and if you still have issues, well... rma it.

     

    Also, if the board STILL misbehaves, get the bios update on the main eVGA site. It very well may help. And whoa... 10,000 3Dmarks? How the hell? I'm fighting for 7400 right now and you're at stock? What the hell? Maybe it's cuz I only got 512MB of RAM in right now.

     

    That dude got SLI. :) I assume we both are non-SLI? I'm getting around 6500 stock. I think I posted my score above.

     

    Are you getting the same error msg I am? After clicking "Don't Send" I get another dialog box that says:

     

    The instruction at "0x7c9122ba" referenced memory at "0x00000000". The memory could not be "read". Click OK to terminate the program.

     

    The value changes though - one would have instructions at "0x7c910f29" referenced memory at "0x00000300". Another time, it would say "write" instead of "read".


  12. Just came back home - it crashed. Says something like "can't write to memory". Other errors says "can't read memory". This is even at stock speeds.

     

    I don't get it. Memory bad? But it passes Memtest86 for 9+ hours. Even Prime95 ran for 9+ hours - and I made sure it was a blended test with around 925 MB for each instance so it would cover most of the memory installed.


  13. It is the latest non-beta build off nVidia's web page.

     

    ForceWare Release 80

    Version: 81.98

    Release Date: December 21, 2005

    WHQL Certified

     

    I cranked the core down to 448 MHz and see if that helps. It is currently running the loop now while I stepped away to visit a friend. i'll check back and let you guys know what happened when I get back.

     

    Am I suppose to go insane now? Should I loose it yet? lol


  14. I run my tests at 1024x768, 4x anti-ailising and anistropic filtering. Your have yours set to optimal and no texture filtering.

     

    When I set it that way, it's fine. It's when I turn up the quality (more antialising and best texture filtering) is when it crashes. It also does it in trilinear filtering, too.

     

    I just DLed coolbits and I'm going to check it out.

×
×
  • Create New...