Jump to content


  • Content Count

  • Joined

  • Last visited

About ShinyG

  • Rank
    New Member
  1. Since I was really bothered about why my new system isn't working, I decided to investigate a bit. I ran memtest with one stick in the first orange slot (marked DDR2) and saw that the sticked passed both 10 passes of Test 5 and 8 passes of Test 8 @1T. I tried running dual-channel @1T again: 2.61V resulted in reboot after "DMI Pool Data bla bla bla" 2.66V resulted in reboot after Win XP 64 bootscreen 2.71V resulted in reboot at login screen 2.77V and higher resulted in reboot right after login So I decided to reflash the bios. This time I used Merlin's EX041106 bios and guess what? IT WORKS! Weee, stability at last! Not it's time to burrrn it . I'll make another thread and keep you posted on how the action goes. Until then, this thread is no longer necessary! Thanks!
  2. First off, sorry for the delay in responding, I took a week off to visit Transilvania . I just got back and had the chance to run Memtest. Here are the results with both sticks in running 1T: 1. With all settings on auto, except CPC, memtest first caused a reboot at Test7 2. With all settings manually set, memtest froze after 11 minutes at Test5 3. With all setting manually set and 2T, memtest successfully completed one pass before I stopped it. Any ideas on what might be wrong? Is it the mainboard or the memory?
  3. Right now I've loaded EX406BTA from Tmod's CD but it still crashes at Windows boot when using both sticks @1T. I've set it to 2T, and not it seems stable enough for me to write this message
  4. Hi, I just recently managed to "blow up" my NF4 Infinity MB and bought a used NF4 eXpert to "compensate" I flashed the 704-2bta BIOS following the instructions but I have major stability problems I did not have with the Infinity. (Note: I had the same problems on the original bios also). I've set all the timings to auto to make sure I'm running stock. Here are the memory configs that are not stable/don't work: 1. 2x1Gb stick using slots 1 and 3 (not recommended in manual) - NO POST 2. 2x1Gb stick using slots 2 and 4 (recommended in manual) @1T - BSOD or Windows freezes shortly after boot. 2. 2x1Gb stick using slots 2 and 4 (recommended in manual) @2T - not stable but usable. Still freezes/restarts from time to time. 3. 1x1Gb in slot 4 (orange one, closest to CPU) - MB memory error (short beep by MB) 4. 1x1Gb in slots 1 or 3 - NO POST (no MB error beeps this time) 5. 1x1Gb in slot 2 @1T - BSOD or Windows freezes shortly after boot. 5. 1x1Gb in slot 2 @2T - stable. I've also tried increasing the RAM voltage to 2.8 but with no success. This is odd! Can anyone help me with this? Thank you
  5. Well, something was about to give, and that was the mainboard! Seems that not all DFI's are created equals... What a crappy piece of useless PCB is it now with those blown capacitors... I was mad at first for ruining it, but I quickly found a replacement on a trading forum for cheap I'm now the proud owner of a barely used, still in warranty DFI NF4 SLI-DR eXpert :cool: This thread can be considered closed. I'll go ahead an create another one for the eXpert and see if I can actually get my Opty to 3GHz like I've been dreaming since I got it!
  6. Thanks for the info man! Good to hear I don't have to backup 300Gb of data for a MB switch Well, I'll back up the most important stuff anyway, just to be sure. I'll post the results soon.
  7. Hi guys, I have two Hitachi HDDs in RAID 0 on my current mainboard: a DFI NF4 SLI Infinity using the native NF4 Raid controller. I'm currently upgrading to a DFI NF4 SLI-DR Expert. Is it possible to keep the RAID array without loosing my data considering the fact that the chipsets used by the two mainboards are the same? Thank you.
  8. I decided to eliminate the temperature reading problems so I tested the monitoring programs one by one in conjunction with OCCT: Conclusions for the NF4 SLI Infinity. All tests done @ 250 HTT, mem 1:1. For OCCT I used the CPU test and 70C as the upper limit for temp. 1. MBM does NOT work! The temp fluctuations are just too big on the CPU reading. 2. SmartGuradian does NOT work! Same problem. 3. coreTemp 0.95 partially works. It gives temp readings for each core but occasionally makes OCCT crash withe "CPU too hot" error, even thought the graphs show temps not going over 55C... 4 Speedfan 4.33 reads one core temp with the "K8" sensor (I take it it's the CPU ) and the other through the ITE monitoring chip. These two temps are consistent with the ones coreTemp reads. The one read through the ITE sensor goes over 50C and I get the "CPU too hot" errror in OCCT. 5. Everest 4. This one seems to work. It reads the two core temp which I've noticed to be similar to what coreTemp and Speedfan read, but there is no fluctuation at all, so OCCT doesn't crash. There are two more CPU readings in Everest: a. "CPU". Seems to be perfectly identical to the "Motherboard" value (!?!) b. "CPU diode". This sits nice and comfy at about 35C during full load (!?!) Which reading from Everest should I trust? Note: the two core temps might seem high. In fact I'm running the Xigmatek at 65% to keep it silent. I'll pump up the rpms when I go higher in HTT.
  9. I decided to rewrite this post as there's no more need for the questions I asked I started having stability issues even at stock speed so after reading a couple of reviews on the net stating that AS5 and the Xigmatek HDT need special attention, I reset the cooler. So I spread the TIM carefully in a thin layer with a surgical rubber glove and put the cooler back on making damn sure I overlap it properly on the CPU. Preliminary results: alost 10C lower on the second core! Seems I did a poor job of installing the cooler the first time Back to getting the old 939 up to speed... I hope I'll find a way to make it at leas boot at 3Ghz!
  10. I totally agree with Angry on the 3dmark issue. Still, I don't consider Crysis to be a fair benchmark either. The game is obviously favorable towards nVidia, which actually contributed to the game's development. Using the same reasoning, I wouldn't consider HL2 a good benchmark either, as ATI is still actively involved with Valve and Softimage (they do XSI, the 3d program used the most by Valve. Best thing is to compare games that don't have crappy logo's like "nVidia, the way it's meant to be played" or "best experienced on ATI graphics". I hate seeing corporate logos in games...
  11. It seems that I'll have to ditch MBM5. The MBM5 thread here at DIY states there are problems with Infinity boards and MBM5 Any other program that is trustworthy? I'm using coretemp 0.95 at the moment, but when the CPU heats up (300HTT or so), the sensors show a difference of over 10C between the cores) PS: got it stable at 275x9 and 1.425V. I'll try to stop fooling around by going straight to extreme clocks and make my way up bit by bit.
  12. The temperatures I mentioned in my earlier posts are from coretemp 0.95. Well, I tried installing and configuring MBM5 (followed the guide here on diy-street) but I had tons of problems: The CPU temp oscillated between two values: -25C and 105C (!?!). the PC speaker was going nuts with beeps and I decided I'm not going to take anymore of this and unistalled it... I tried speedeuphoria's tip about the async latency and read preamble but Orthos failed miserably... again!
  13. Thanks for the tips on the async latencies. I'll try them as soon as I get home. I'll post some mbm5 temps also. I just realized that I can't remove the IHS on the CPU. The reason is the new cooler I got, which has its 3 heatpipes in direct contact with the IHS. Removing the IHS will only leave the center heatpipe in direct contact with the core, the other two will not be touching anything (
  14. Hi guys, It's been awhile since I posted here. I was busy finishing my university studies and thank God, I'm done. Just the final exam and then I'm going to be a Physics Engineer with a diploma in Laser Technologies. Weee, totally useless in Romania Anyway, this means I finally have some time go get back to toasting the Opty... Since last time I was asking around for advice I did some changes to my rig: 1. I got a Sapphire 3870 after a month of hunting it down through the gray forest that is the Romanian online IT retail sector. The 8800GT was impossible to find in the Holiday Season. 2. Ditched the A64 Pro for a bigger cooler, the Xigmatek HDT-S1283. Big and somewhat silent... Last time I tried, I managed to run Orthos' CPU test for 8 hours @2700Mhz, so I was under the impression it was stable, yet games kept crashing and crashing after no more than 20 minutes of playing, so I was forced to go back to 250HTT and 1:1 divider. Now that I got the big cooler, I went back to 2700Mhz, same settings as before: HTT 300, HTT multi 3x, 1.6V cpu Memory: 166 divider, 2.7V Chipset: 1.6V Orthos' CPU test managed to run for about 10 seconds after which the system crashed. I lowered the HTT to 290 and tried again. This time Orthos worked for 10 seconds, then errors started to appear. I went even lower to 285 HTT. This time Orthos seemed to work well but I stopped it: one core was at 51C and one at 72C (!?!?!) I think one core is acting up and that might cause the instabilities I get. Now the big question: Should I remove the IHS? Will "going naked" mean more even core temps?
  15. It's on Arctic Silver 5 now and stable again I will definitely go for better cooling. I'm thinking of a Thermalright Ultra 120 Extreme for the cpu and some copper sinks for the memory chips , but I can't make up my mind on a decent fan for it... Any ideas for the fans guys?
  • Create New...