View Full Version : DFI Ultra-D Inconsistent Performance / Overclock
dnottis
04-09-2005, 01:15 PM
Just want to see who else has inconsistent results with their DFI Ultra-D. One minute I'm 110% stable at 302 FSB (x9 1:1 1:T), the next minute I cant do 2.6 Ghz. I've tried different ram, PSU, CPU, reseated CPU Cooler & Chipset cooler with As5. Basically I'm done with this board its just not reliable.
flexy
04-10-2005, 12:47 PM
define "stable" ? If you run OCCT (and i really prefer OCCT over Prime)....say set memory to test at 750MB [like i do]...and it passes some hours.... then you want to tell me the next time you run it at the same settings it would fail ?
Ever considered your problems might be related to the programs you run (what are you using for stability testing ?)....or something else interferes with your testing.
i have exact same inconsistency with my SLi-D so i've clicked yes, some days 2.5ghz+ is fine, ie. a good few hours gaming is what i call stable, yet others it wont do 2,2ghz without bsod'ing, think i'm heading over to msi's ranch again
dnottis
04-10-2005, 01:57 PM
I ordered the MSI Neo4-F - I'll report on this week. My DFI stopped seeing USB and is now hanging at the device ID screen again... argh I hate this board.
Minstadave
04-10-2005, 02:07 PM
No probs here :confused:
dnottis
04-10-2005, 03:29 PM
Personally I think it's a lemon. One day I can run 305 1:1 1T x 9 @ 1.53, the next day 290 it's stable, 12 hours later 305 is stable again. I've swapped everything else, CPU, PSU, memory, moved all devices to PS2, still it does the same thing. Another issue I have is failure to load, freezes at the device ID screen after post and the USB devices not working frequently.
All in all this has been the worst motherboard I've used for issues, stability, reliability - Im sure it's just a bad board, seems there are lots of them out there, high DOA, RMA rate I'm still kinda sour to DFI now.
GregP24
04-10-2005, 03:54 PM
I've been running my Ultra-D since Friday night, and I'm having NOTHING but problems. 4 out of 5 times I can't boot into Windows without it BSODing and rebooting.
With the exception of the XL, everything is just as it was on my Neo2. I thought the vid card might be borked, but I swapped in my 6600GT and the problem remained.
Everywhere I go, I hear about how Jesus himself must have made the Ultra-D. It's godlike among mobos, but I get it and its a POS.
Didn't vote because I have a SLI-DR... but it is extremely consistant for me. I have known good "reference points" I use when tweaking, and the OCability has always been the same for a given set of known good settings and hardware config... as in +/- ONE Mhz... and usually spot on.
Kosmic71
04-10-2005, 04:57 PM
No problem here, but I've got a SLI-DR, same board anyway.
joe2004
04-10-2005, 05:53 PM
I have two Ultra-D boards. One is doing that exactly as described, one day 308, another won't do 302. Just plain irritating.
The other board does OK. I checked two different CPUs and the same thing.
Some of these DFI boards are crap really. Very inconsistent manufacturer.
I use to be a big fan of Abit and Epox but unfortunately Abit is now making expensive gadgets and Epox just can't get their boards in order.
flexy
04-10-2005, 06:40 PM
wrong/old/beta bios ???
bios 310p here and running fine.
also, OP, you didnt answer my question. Another possible explanation would be that your sys is not really stable in the first place.
my guesses:
1) bios (also including weak bios/cmos battery ????)
2) voltages in bios ? (wrong settings ?)
3) not really stable because you used the wrong programs to test stability
4) board actually IS a lemon
dnottis
04-10-2005, 07:12 PM
my guesses:
1) bios (also including weak bios/cmos battery ????) I've tried 5 bioses from 125 to 218 to 310, 310P and 331-3 (for TCCD)
2) voltages in bios ? (wrong settings ?) Nah, they are fine.
3) not really stable because you used the wrong programs to test stability - the same programs and actual exes I use on my Neo2 platinum, so thats doubtful
4) board actually IS a lemon[/QUOTE] YES! DING DING DING!!
flexy
04-10-2005, 07:15 PM
dnottis, and others.....
my experience with the board showed me that the board itself is fine, stable and everything - BUT CAN BE EXTREMELY TROUBLESOME after certain tweak operations/overclocking sessions, playing with timings etc.
I dont know how else to explain this....right now i know if i leave my board at this overclock and do NOT "experiment"
( see http://www.xtremesystems.org/forums/showthread.php?t=59089 )
i will have ZERO problems whatsoever.
As soon as i get in a tweaking mood i might run into trouble, and as you see in above thread i am not even talking about tweaking on WIndows XP level, but eg. solely in BIOS /w memtest.
The PROBLEM with this board is, that, in comparison to my former sys, eg. AMD XP1900 THIS board and CPU has so much more features, and BINGO! i also have a LOT more hardwae-level tools running and tweaks and hardcore tweak utilities and all this stuff...well it does NOT contribute to more stability.
--> With AMD64 and CnQ we have voltages and CPU multis possible to change ON THE FLY in WIndows
--> A64tweaker, changes CAS latencies and as good as everything else on the fly.
--> clockgen...etc..etc...
--> constant overclocking and playing around with settings, which i HARDLY did with my former system.
I remember once i figured the old system ran at 140FSB...well i set it, tested it with prime and memtest....and then forgot it and did NOT tweak it again ever since i knew i reached my limit.
The fact that you write "one day it works with HTT XYZ the other it doesnt" means that you're still tweaking around....and as said, there are some bizarre issues with this board and tweaking CAN seriously screw up the whole system even if you THINK you are stable (eg. memtest) etc.
I am not talking out of my a** but from experience and i can list some of the issues i see with this board alreadu - and never had SIMILIAR issues with my former Dragon+ board:
1) the most fault-sensitive moment sems to be while boot, before the logon screen.
What i saw there already:
BSODS, Hangs, Page faults
something going on with memory manager, i dont know.
2) file corruption of system files in windows/system32/ eg. certain *.sys or *.dll files can be damaged, DONT EVEN ASK ME HOW its possible, but i verified it with the signature verification checker. Files get overwritten and contain only zeros etc. need to be restored by deleting them so windows will restore them automatically.
Usually happens after some heavy tweaking, i now have a routine of checking all my system files after i had heavy crashes eg. while (stupid me) i use tools like A64tweaker etc.
OK what i want to say is:
Board is able to run nice and stable - but touch it and tweak around, and be prepared for headaches.
Also...i am just back from one or two days with EXTREME issues on this board because i was stupid and installed a creative audigy beta driver. The driver was probably not intended for my audigy2 value (different chip)...anyway i tried a hard way and somehow got the drivers installed. No problems *expected* whatsoever.
Well...i spent the following days cleaning my whole system with half a dozen tools (regleaner etc., drivercleaner)...and STILL had a hang while boot which did not allow me to enter XP except in safe mode.
tried several system restores, several attempts literally wiping out EVERY trace off creative drivers...editing/cleaning registry, using windoctor, drivercleaner etc. :)
Well i dont know but after one day i somehow got it ok now, re-installed my normal drivers off the CD.....and i am finally able again to boot into my windows and everything is stable and great as nothing ever happened.
This maybe just as a reminder that this board can be be major headache - and if i just leave it alone as it is right now and overcome my urge to tweak and install beta drivers etc...then i also know i wont have problems with it :)
ben805
04-10-2005, 09:23 PM
I have two Ultra-D sitting infront of me now, the first one I bought about a month ago, the raid on that one simply refused to work regardless what bios or settings I use, the second Ultra-D is only couple days old, this one can recognize the raid but winxp or 2k3 server refused to recognize the array, I have been spending the whole darn weekend trying to get the raid working so yea I have tried every single things out there, swapped SATA cables, different rail voltage, ALL bioses, different SATA ports, different drivers and diskettes, you named it I've done it.
Now when it comes to consistency, the first mobo can do 326 x8 all day long and cpu only require 1.45v, the 2nd mobo can only do 325 x8 but require 1.5v, for whatever reason all voltage reading on the first mobo is much lower than the second one, voltage regulator seems to be different on both motherboard. Though the cpu temperature is lower on the 2nd mobo and I have no idea why it require more voltage than mobo 1 to run at the same speed either!! I'm not about to go get the 3rd motherboard, two attempts are enough for me and I'm done with it as well, can't deal with this 'love and hate' feeling I have toward these ultra-D anymore :slapass: so I'm going to return both motherboards if possible, if not then I'll just keep the 1st one as backup and return the 2nd mobo, going to give MSI Neo4 a try as well :D
mitgsx
04-11-2005, 05:14 AM
How about this, I get my system stable after days of fighting it. Ran stable for 2 days, priming over night. Third night no prime but wake up to the comp just hanging on the bios splash screen. Had to clear cmos just to get into the bios (310 oem). Would not run anything close to them settings again.
I have tried 2 different sets of ram also, sig below and geil pc4000 2 x 512.
I am about to sell off this system and get a AXP setup again. I am tired of gaming against my brother-in-law who owns my old T-bred 2600+@2.4 daily, He plays and I keep :banana: :banana: :banana: :banana: ing rebooting. This is sometimes at stock settings too.
I have had it prime blend, 3dmark-01-03-05, occt, S&M, super pie 32m stable at 2.45, then the following days I would wake up to it hanging. The ram doesn't run consistant also. Some days it will run higher than other days, this is testing with memtest 1.51.
TiTON
04-16-2006, 12:57 PM
** Sorry.. DIdn't know I revied an old thread over a year ago.. I just looked at the month and date.. didn't pay attention to the year. My Bad :( **
I was searching the forums to find my help with my Ultra D inconsistent OC performance.
I was playing with my 0601 x2 3800 on DICE (cpu has no cold bug). I would trying to find the max CPU speed (I factored out RAM and LDT).. On my 1st try I got like 3.38Ghz on only 1.55vCore before it hung.
Then I rebooted the machine, and figure I will try to run SuperPi and such at 3.3Ghz to see what I can get.
THe moment I kicked the OC to 3.3 it hangs.. I was very frustated.
So I rebooted, and took baby steps again.. this time my pc hung at less then 3.2 Ghz.
I mean.. how can a Max OC go from 3.38 to not even 3.2 Ghz?
As I mentioned, I factored out RAM / LDT, to focus only on CPU. And with DICE cooling, I know the CPU was running just fine.
Any suggestion on how to get my OC stable?
SAD
BioPC
04-16-2006, 03:26 PM
i owned 2 Ultra-D. Both RMA'ed.
First one corrupted BIOS, second one sata raid was random working.
That's indeed the most problematic mobo i ever had.
Now i'm happy with a much cheaper Asrock ULi till i find something i can trust beter.
Repoman
04-16-2006, 03:34 PM
lol, good use of search function though!
babalouj
04-16-2006, 03:55 PM
both my Ultra-D's have consistancy problems (one is one of the first revisions and the other is AA0). Some days it just wont boot into windows at 2.7ghz (my 24/7 settings) and other days it will prime 24 hours at 2.9ghz. The other board just wont boot above 270fsb ever no matter what chip or memory.
ahmad
04-16-2006, 04:01 PM
Never had this. I can always wake up in the morning change all the timings and settings to what they need to be and it boots right up. Change again to something I know works and it has never failed. Thats with a couple of Ultra-Ds as well not just one...
Which BIOS(es) have you tried?
@_dud
04-16-2006, 06:02 PM
Mine runs fine just fine, and I only had a problem once with stability, then I recet the CMOS and it has been fine ever since. Nicest board I've ever owned. :D
HelloIDistance
04-16-2006, 06:05 PM
I never had any problems with mine. Mines RevAD0 been running fine so far at 2.83Ghz stock volts, 24 hour prime stable.
IvanAndreevich
04-16-2006, 06:09 PM
I had built probably close to 10 systems (or went through that many boards anyway) with DFI NF4 boards (SLI-DR, Ultra-D's) and none had that issue.
AB0 and AD0 both cannonball solid. (well, I drowned the AB0)
afireinside
04-16-2006, 07:02 PM
My AD0 runs perfectly 24/7 no issues, AB0 ran CH UTT to 267, HTT to 360 @ -60c, 2.05vcore, 4vdimm, 3.6ghz, all without a hitch and never an inconsistancy. Both ran 6-23 BIOS.
Hassan
04-16-2006, 07:52 PM
My 2 ultra-d's run fine... only inconsistencies I find is every few weeks I teak more and improve my mem timings or oc.
TiTON
04-16-2006, 09:35 PM
lol, good use of search function though!
Hehe.. thanks repoman :)
Well.. since there are lot of people that seem like they want to help me out, I will go ahead and see if i can pick your brains :)
I have primarily been an Intel man.. this is my first DFI Utlra D (second hand). I already modded it to run SLI, which is great.
Now here is the question I have:
1. Where is this revision code at? Is one better then the other?
2. Which Bios should I use for my X2 and the following mem:
2a. Corsair XMS Pro 3500LL (2x1gb), which are Infineon C.
2b. Geil One S 0519 TCCD (2x512mb)
Do you think the board is faulty? Or maybe the BIOS?
SAD
Ku-sama
04-16-2006, 09:38 PM
i wouldent know yet, my SLI DR-Expert came DoA (kinda, it wouldent let me install an OS) so ill report back soon
Hassan
04-16-2006, 09:47 PM
Have you tried the newest bios? It is stable for me and is supposed to be good for multiple dimms and 2GB clocking, it really helped get the most out of my new ram
shimmishim
04-17-2006, 11:03 AM
i've had my ultra-d since january of 2005 and it's still alive and kickin'!
i've been thinking of upgrading to an AD0 rev board though.
ozzimark
04-17-2006, 11:19 AM
mine's pretty predicable.. from board to board though, there's a huge variance in results. :stick:
@_dud
04-17-2006, 03:25 PM
Hehe.. thanks repoman :)
1. Where is this revision code at? Is one better then the other?
2. Which Bios should I use for my X2 and the following mem:
2a. Corsair XMS Pro 3500LL (2x1gb), which are Infineon C.
2b. Geil One S 0519 TCCD (2x512mb)
Do you think the board is faulty? Or maybe the BIOS?
SAD
1.Rev. code is on the side of a PCI or PCI-X slot.
2.I haven't changed bios so I don't know
I think the bios is faulty, but that's just a guess.
One_Hertz
04-17-2006, 03:57 PM
My ultra-d is probably the crappiest board I have ever owned. the oc is ridiculously different every day, up to a difference of like 300mhz on my 3800+ X2... Setting up raid was freaking hard too, it would bsod on me every time I tired to install raid drivers during the installation of windows. I had lots of similar problems and it took me about 24 hours of work overall to get windows on it because of problems like the raid one. I also have an x-fi and everyone knows how much of a pain that is. The stupidest part is that to fix half of the problems all I had to do is switch off the power and unplug and plug back some connectors. If thats not inconsistency I dont know what is.
The worst part is that I cant even rma it because it runs ok on stock and I already fixed all the other problems. It pissed me off enough however for me to just take a baseball bat to it, so I'll probbaly soon do that and get myself a nice r580 board.
Edit: Forgot to mention that it takes a good 10 seconds to post and sometimes it freezes during post. Even saw display corruption during boot that resulted in a freeze as well.
WeStSiDePLaYa
04-17-2006, 03:57 PM
check components for dry joints. seems its becoming more and more of a problem as companys switch to lead free solder that needs to be heated more to melt.
S0nic
04-17-2006, 06:09 PM
I've had my ultra-d since Jan last year and spent quite a bit of time overclocking and tweaking. I sometimes get it stable (memtest, prime, occt, super pi, 3dmark) then the next week it starts acting funny, sometimes it stays stable at a certain speed for a long time (month or more) then craps out again. My temps are reasonable so I don't that's what's causing the problem.
I've been thinking of switching to an AD0 since those seem to be a little more stable from what I've been reading here, or I might go for a neo4..
vBulletin® v3.7.0, Copyright ©2000-2008, Jelsoft Enterprises Ltd.