DFI UT X58-T3EH8 continued - Part 2

[Praz]

1: I am getting a lot of lockups on CF and then lots of reboots and recycling back to CF from code 26.... if I make it past 26 then everything is fine... (quite often it is 3 or 4 cycles - cheers for the info Praz)...11 cycles?

DFI has been able to duplicate this issue and are currently working on a fix.

For anyone who is having major problems with the D0 and Elpida at 200x20/2000 memory there is a new ABS file posted. It is for use with the 02/17 BIOS as it has proved to be the best version at getting around the issues users are experiencing.

http://www.dfi.com.tw/portal/CM/cmpr...rue&itemId=561

Like most of the ABS files the voltages are quite a bit higher then most of us use. VTT is set to 1.50V. My board will boot into Windows at that setting but will not pass any stability tests. 1.36V - 1.42V will probably be where most users will end up, assuming an average D0 CPU. After finding the optimal VTT voltage the other voltages can be adjusted while fine tuning VTT.

Boot UP Clock is set to 135. Can try 130 - 160 with 135 - 145 being optimal for most setups.

There will be a BIOS released based on the 05/08 that addresses these issues.

[PhenomII2]

Cheers guys .. gotta wait 2 weeks for this baby to get it @ my place , but i guess i ll wait now

[Sam_oslo]

Play around to find a sweet spot, and 19x seams to require less Vcore than 20x, just like my good old i7 920 CO., But i couldn't dream av primestability at BCLK 220 with C0.

VVT: 1.24/1.28v (BIOS/Load)
NB Core: 1.13/1.12v

w3520 @4.18GHZ(220 x19), 1.2v , 1 hour primestable


I have to test more before i go further with this setup. I don't know how my GTX 295 would react to BCLK 220. So far so good and if this setup "behaves", i may use this for 24/7.

[ci2kla]

What the hell is this?!

[ChaosMinionX]

maybe the acid smell was the paint job?

I cannot even get the boards power buttons to turn the board off after I had that acid smell + code 88.... it requires that I cut the power the PSU.

I tried several memory configurations as well.

Any thoughts?

Edit: no matter what I do... Clear cmos or any combination of memory in dimm, it simply will not boot. It sticks at code 88 and I get no further.

Is an acid smell normal on dfi when starting it up? I haven't used a dfi since the RD600 days.

[The Water Dog]

What the hell is this?!

Vcore i think, i use my UD4P board now...but after what i remember that is vcore.

[ecidious]

Vcore i think, i use my UD4P board now...but after what i remember that is vcore.

He's questioning the screwy order/double appearance of the 1.06255V entry.


Has anyone else noticed that their board is very sensitive to settings while changing GPU's and such? Mine needs more vCore for every additional card, something I thought the aux power connectors were supposed to help prevent.
Switching from an 8800 GTS to a GTX 285 requires 0.02 extra vcore as does adding my raid card. PSU is an Antec Signature 850 and the rails are rock stable.

[kitfit1]

Has anyone else noticed that their board is very sensitive to settings while changing GPU's and such?

Nothing unusual about that at all really. All mobo's tend to react in the same way when changing GPU's. Even more so when changing to a more powerfull GPU then was installed before.

[MomijiTMO]

I cannot even get the boards power buttons to turn the board off after I had that acid smell + code 88.... it requires that I cut the power the PSU.

I tried several memory configurations as well.

Any thoughts?

Edit: no matter what I do... Clear cmos or any combination of memory in dimm, it simply will not boot. It sticks at code 88 and I get no further.

Is an acid smell normal on dfi when starting it up? I haven't used a dfi since the RD600 days.

You know I was joking right

Any electrical burn/acid smell is not normal and something let go on the board. If you follow the steps in Eva's guide and the board doesn't get going, it could be dead.

[The Water Dog]

He's questioning the screwy order/double appearance of the 1.06255V entry.


Has anyone else noticed that their board is very sensitive to settings while changing GPU's and such? Mine needs more vCore for every additional card, something I thought the aux power connectors were supposed to help prevent.
Switching from an 8800 GTS to a GTX 285 requires 0.02 extra vcore as does adding my raid card. PSU is an Antec Signature 850 and the rails are rock stable.

bah, i saw it now. Stupid me...sorry:S
I tought it was strange that he was asking about that:P
But making sense now.

[ecidious]

Nothing unusual about that at all really. All mobo's tend to react in the same way when changing GPU's. Even more so when changing to a more powerfull GPU then was installed before.

I was just wondering whats up because none of the last 3 systems I had before this one with the DFI needed any changes when I put the card in.
Maybe I've just been lucky

[p2501]

Okay, before I went to bed last night I had 210x19 chugging along with 10 LinX cycles. Pwr on today in the morning and I find Vista isn't booting anymore (goes until you see that bar 'Windows is loading', I could hear the hdd being accessed, but then it stops and stays there), figured maybe the BIOS is corrupted so flashed 05/08 over 02/17. Still not booting, so maybe Vista is shot. Started reinstalling and guess what, it seems my raptor can't be found using AHCI mode anymore, IDE is fine and I've got a working install again. Didn't try another hd. Any ideas, did I just shoot my controller or HDD?

[Sam_oslo]

I have been running @4.01GHZ(19x211) for 24/7 use on both i7 CO, and w3520 and never had any problems. I have had a few short (10-20) Linx run with SSD on SCSI-mode too.

I have got a few IRQ_xxx errors on USB devises while priming at 220 BCLK, but nothing is gone bad yet. High BCLK will push the SB and PCIe more, and probably Linx has shoot out your AHCI controller.

EDIT:
How much volt on SB did you have?

[p2501]

Bone stock on the SB. I'll definitely test this out with another harddisk, god that would suck if you're right.

Edit: Controller seems to be okay. Another disk installs just fine on the same port so I guess the raptor is faulty. I had it running on a cardboard box without extra cooling. Fun fun. Time for SSD maybe?

[Sam_oslo]

Good news, and yes you should go for SSD. I have been using a OCZ Vertex 30GB as OS disk (W7 RC 64bit) and can never think of going back to HDD again. It is cool, fast and quiet.

[ChaosMinionX]

So I just decided to RMA the board, nothing would get it going just code 88 everytime... better luck with the new one I hope.

[p2501]

So I just decided to RMA the board, nothing would get it going just code 88 everytime... better luck with the new one I hope.

Yup, good luck! Acidic smell and boot issues for a new board, maybe a dud? I hope you get a new one fast.

Good news, and yes you should go for SSD. I have been using a OCZ Vertex 30GB as OS disk (W7 RC 64bit) and can never think of going back to HDD again. It is cool, fast and quiet.

Yeah, I'm still unsure if a Vertex or that 80Gig Intel SSD would be better. I guess when going from HDD to SSD both brands are fine.

On my side of things: for stabilizing bclk 210 I gave IOH/ICH 1.1V Voltage and IOH Analog Voltage each .1v more juice, is this within sane limits? What is this refering to anyway? To a part of UnCore or the Northbridge?

Edit: There seems to be a direct correlation between IOH volts and Linpack times. Going from 1.15 on both to 1.17/1.5 adds nearly 5 seconds to completion time of EACH run! More volts, less GFlops.

[emertX]

Is there any new beta BIOS on track?

[cstkl1]

will post back later prime 12 hours stable again

difference now is
vdimm at 1.58v for 2k C7
and vcore is now same as linx 100 pass at 1.24

ok here is what u do

set dram pwm to nominal
set vtt pwm to reduce 15%

set dqvref to +1.5%

also another note

cpu vtt add disable and vdropp disable

u either set both on enable or both disable.

[StimulatedBoy]

Is there any new beta BIOS on track?

Beta BIOS 2009-06-16 now available on DFI site but only as a Diamond Flash Image. Now also added the "useful" one...

[ci2kla]

Beta BIOS 2009-06-16 now available on DFI site but only as a Diamond Flash Image so far.

Time to try some more new things!

[OUT_TOPIC]
The Heatkiller 3.0 is all out of stock both in Aquatuning, PerformancePCs, Sidewinders, Chilledpc OMG...except one place that still have 6 pcs in stock...Just ordered two
[/OUT_TOPIC]

[ci2kla]

A Quick Test.

This XEON simply ROCKZ!~

[pangingIII]

just got this board and set it up a few days ago, i have already tried bios 0217, 0428 and 0508.
but i don't have a lot of time with it yet, also i'm still waiting for better ram kits, as of now i'm using crucial PC3-10600 1333MHz 9-9-9-24 Micron D9JNM/D9KPT
anyway, i like to thank eva2000, for a lot of useful informations. i'll continue checking most of your posts to learn a lot of things you've found about this board.

here are some of my quick tests which most of them using AUTO + VID special add vcore.








4.2GHz (Turbo mode enabled), full load 1.248v (cpuz), no load 0.976v (cpuz). room temp. 27c, True+2fans.

cpu idle temp...


running prime95...



4.3GHz Superpi 32M Win 7


4.4GHz (Turbo mode enabled) Superpi 32M Vista x64

[MacClipper]

Not sure if anyone has posted this already but brand new BIOS was just announced by RGone,
http://csd.dficlub.org/forum/showthread.php?t=10674



PS: Thought it interesting to note that the UT has the smallest size BIOS update at 527KB and the Jr the largest at 559KB.

[ci2kla]

A Little PI 32M Run~