DFI UT X58-T3EH8 continued - Part 2

[ChaosMinionX]

ChaosMinionX which bios do u use? Have you tried running Linx on these settings?

6/19 BIOS, will post LinX 100 passes when done.

[p2501]

Out of pure curiosity, what are the max temps you guys get when running LinX? I have the very same cooling setup that Eva has (taken from my trusty old Q6600), but under LinX I'm still reaching 88°C with a difference of 8°C between hottest and coldest core. Is that a sign of bad solder-die or solder-IHS contact? Or is this completely normal for these suckers? I lapped it to a mirror, AS5 still hasn't cured and until now I don't see a difference.

[rioja]

ChaosMinionX
Did you try your Corsair Dominator 1600C8D at higher speeds? I can't run them even at 1640 8-8-8-24

[ChaosAD]

6/19 BIOS, will post LinX 100 passes when done.

I ll try your settings on 6/19 and i ll report back, 100 passes of Linx is too much imo, 10-15 passes is fine for me to call it stable

Out of pure curiosity, what are the max temps you guys get when running LinX? I have the very same cooling setup that Eva has (taken from my trusty old Q6600), but under LinX I'm still reaching 88°C with a difference of 8°C between hottest and coldest core. Is that a sign of bad solder-die or solder-IHS contact? Or is this completely normal for these suckers? I lapped it to a mirror, AS5 still hasn't cured and until now I don't see a difference.

Max temps i ever saw was under Linx was at 21x200 HT Off, 1.39vcore, ~25C ambient with WC in my sig, that was 68-65-65-61.

[ci2kla]

sorry OT...I'm worrying about my GTX 275 as it went crazy temp when furmarking....default voltage 1.125v under load can hit 105C within 2 minutes ambient was 25C...........please help...

[p2501]

Max temps i ever saw was under Linx was at 21x200 HT Off, 1.39vcore, ~25C ambient with WC in my sig, that was 68-65-65-61.

Okay, 7 deg. difference between hottest and coldest, but on a totally different level. My watercooling setup is in its final stages of purchase so I hope I'll have those temps soon too. I can't even Prime because if it should BSOD then I don't know if it's the heat or something else. Besides, 13h+on 86°C can't be good. Thanks for your input.

sorry OT...I'm worrying about my GTX 275 as it went crazy temp when furmarking....default voltage 1.125v under load can hit 105C within 2 minutes ambient was 25C...........please help...

Take the HSF off and check if it has bad contact/change TIM with something real would be my advice, but I guess your card is new right? Do you reach those temps with fan nailed on full blast (get rid of Auto setting)? If you're still that high, maybe it's time to revise case air flow? And lastly, nV had a habit of putting chips out on which the heatspreader had extremely bad contact with the core, but I didn't know that this is the case with 275 type of cards.. in that case -> RMA. In addition, around 105°C it should normally shutdown.

[Biker]

Anyone got this board to work with an LSI card yet?

No joy here in any slot...

Cards are known good (8708ELP + Fujitsu siemens S26361-F3257-L256) and both are using the most recent LSI firmware....

The board is currently using the 06/16 bios.

Failing that if anyone has access to the bios team my current DFI contact is blocking access atm and I need acknowledgement that the bios team is aware or parhaps working on this issue... if you do or you have a newer board bios then please PM me as I need to get testing done etc and confirm problem is fixed or it is RMA time in a week and I do not want to do that...


ps. Many other cards (adaptec 3ware etc) have been tested and are working fine but i need the LSI cards to work with this board.

[ci2kla]

Okay, 7 deg. difference between hottest and coldest, but on a totally different level. My watercooling setup is in its final stages of purchase so I hope I'll have those temps soon too. I can't even Prime because if it should BSOD then I don't know if it's the heat or something else. Besides, 13h+on 86°C can't be good. Thanks for your input.



Take the HSF off and check if it has bad contact/change TIM with something real would be my advice, but I guess your card is new right? Do you reach those temps with fan nailed on full blast (get rid of Auto setting)? If you're still that high, maybe it's time to revise case air flow? And lastly, nV had a habit of putting chips out on which the heatspreader had extremely bad contact with the core, but I didn't know that this is the case with 275 type of cards.. in that case -> RMA. In addition, around 105°C it should normally shutdown.

Yes already checked and re-append paste but still no joy...Tweaking voltage VDDC to 0.9875v will only mae things slightly better...around 73C under load. That's bad...anyway I'm planning to make an GPU Core only WC so maybe I won't care...

[ChaosMinionX]

ChaosMinionX
Did you try your Corsair Dominator 1600C8D at higher speeds? I can't run them even at 1640 8-8-8-24

Unfortunately the 1600C8D have very little headroom, they seem to drop off regardless of voltages around where you are getting.

They also will not do C7 @ 1600.

Storm last night knocked out my power, so I am redoing the test.

[Peen]

Im at 1524 8-8-8-19 on my OCZ it won't go any higher its stable here but wont really boot any higher. weird.

Something is fishy with this board and ATI cards

[tobe22]

I'm currently running 12 GB's of Corsair TR3X6G 1600C8D, on the mATX DFI X58 H6, at 7-8-7-20-1N, ran LinX 75 passes last night at 1.603v in bios (guardian reports 1.58v), all other voltage settings default; one set of 3x2 from Amazon, one set from newegg.

i920 D0, 200 x 18 (turbo OFF) on air, boots at 19, and 20 multi, but a high PWM temp (90+) is keeping me from completing LinX -- gonna try an H8 when I get some time (and newegg ships it for free =P) and put a fan on the cpu area; a 120 in the cpu area isn't help much with the mATX board.

These are the options I'm currently running (when I test 19 and 20 multi, VR Max is Disabled, and Thermal Control is Enabled):

CPU Feature
Set VR Current Limit Max: Enabled
Thermal Management Control: Disabled
EIST Function: Disabled
CxE Function: Disabled
Execute Disable Bit: Disabled
Virtualization Technology: Disabled

[rioja]

i have a little problem, i use following settings

CPU Feature
Set VR Current Limit Max: Enabled
Thermal Management Control: Disabled
EIST Function: Enabled
CxE Function: Disabled
Execute Disable Bit: Enabled
Virtualization Technology: Enabled

Vista profile set to High performance.

And I set enabled x21 multiplier, at idle programms report this 21.
But then i start a load, multiplier get down to 20 immediatly..
what's wrong?

but a high PWM temp (90+) is keeping me from completing LinX -- gonna try an H8 when I get some time (and newegg ships it for free =P) and put a fan on the cpu area; a 120 in the cpu area isn't help much with the mATX board.

Yes, this digital power makes DFI hottest board on x58 as I see. Even EVGA run cooler.
I use waterblock on mosfets, and system temp stll goes upto 70C without direct airflow.

[ChaosAD]

Yes, this digital power makes DFI hottest board on x58 as I see. Even EVGA run cooler.
I use waterblock on mosfets, and system temp stll goes upto 70C without direct airflow.

Then you must have some kind of cooling issue. My PWM at 4.2Ghz get to 68C max with no airflow at all (NB is WCooled). 70C by using mosfet WC seems a little too much.

[ChaosMinionX]

i have a little problem, i use following settings


Vista profile set to High performance.

And I set enabled x21 multiplier, at idle programms report this 21.
But then i start a load, multiplier get down to 20 immediatly..
what's wrong?



Yes, this digital power makes DFI hottest board on x58 as I see. Even EVGA run cooler.
I use waterblock on mosfets, and system temp stll goes upto 70C without direct airflow.

Something wrong with your PWM temp, what is ambient? My PWM sits around 50C idle never above 65C at LinX or Prime95 Load, ambient temp is around 29C.

[rioja]

Are you with HT on or off?
I reinstalled mosfet block 2 times, and got similar results, tried it at open stand and in case.
Now at 4200 1.32v HT ON system temp is near 63C under Linpack. Prime gives a bit lower temps.
But this block allowed me to lower system temp at 15-20C comparing with MCW-30 on NB (before I may reach +85C easy).
And 70C I'm getting if run 210x20 or x22 1.4v, but those jumpimg multiplier 20-21 makes things unclear.

Now I'm stable with 200x21 1.32v 1600 8-8-8-24, but now those 1600C8D is limiting becoz they don't want to go above 200x8 with CL8.


I'm going to try 1640 or 1680 with CL9
Ambient 25C, water 33C under Prime load.

[ChaosMinionX]

HT on, very strange... you arent running PWM Frequency increased are you?

What voltage settings for PWM are you using?

[rioja]

Nope, I didnt change anything in voltage section except VTT, Vcore and Vdimm. Those area between mosfets and CPU, where many small details, is very hot.

At idle 4200 1.32v system temp 43C.

[Praz]

rioja

If you are using 3x2GB Corsair Dominator 1600C8D why is CPU-Z showing 4GB?

[loonym]

rioja

If you are using 3x2GB Corsair Dominator 1600C8D why is CPU-Z showing 4GB?

My board occasionally does the same. I remove the sticks, rotate them, and I'm back to 6g

[rioja]

rioja

If you are using 3x2GB Corsair Dominator 1600C8D why is CPU-Z showing 4GB?

Yes, it is 1600C8D but I use only 2 modules for awhile

[loonym]

Yes, it is 1600C8D but I use only 2 modules for awhile

Well there, that explains yours. Not sure about mine

[Sam_oslo]

@4.2GHZ would make PWM a few degree hotter, even with low Vcore and low VVT.
You need good/top cooling to keep the heat nuder the control in these speeds, both on CPU and NB!. Higher NB-core and VVT-volt would make it even hotter, too.

Here is my @4.2GHZ with 1.23v/1.24v (Vcore/VVT) with a strong fan (Intel's stock CPU-fan) blowing directly on NB at 2039RPM, and that's all it takes to keep a comfy 55'C system-temp on my setup :

[Praz]

My board occasionally does the same. I remove the sticks, rotate them, and I'm back to 6g

Try a different CPU. If it still happens RMA the board.

[eva2000]

My board occasionally does the same. I remove the sticks, rotate them, and I'm back to 6g

tip #16 http://i4memory.com/102537-post3.html

[rioja]

@4.2GHZ would make PWM a few degree hotter, even with low Vcore and low VVT.
You need good/top cooling to keep the heat nuder the control in these speeds, both on CPU and NB!. Higher NB-core and VVT-volt would make it even hotter, too.

Your cpu is very good if run 4200 1.24v
I have Heatkiller LT on CPU, MCW30 on NB, and MIPS on the mosfets.
At 4.2GHz 1.32v/1.34v Vcore/VTT temp stabilized at 62C PWM, 40C NB, 65C cores with 25C ambient under Prime blend. This is quite comfortable temps, but I'd like to overclock more