**Official DFI LanParty UT P35-T2R Review/Overclock/Guide Thread**

[abo333]

ah i see. ok getting now. thanks again George.

edit: how do you run it? when i try to run setfsb.exe I get "sysinfor(x64).sys Error" lol what a n00b.

Hmm..., Please wait for a few days. Author

[Brother Esau]

No squealing here

[ic3m4n2005]

hmm thx for the info so far, i`ve got the Decathlon 750w could this be the problem ? Same sqeeling occurs with DFI 680i but not as loud.

[Brother Esau]

Enermax and DFI have not been a compatible duo for as long as I ca remember.

[retrospooty]

I am on my 3rd DFI board, and always used an Enermax PS with no issues.

dfi lp ut nf4 sli-d and Enermax 535watt - no issues
DFI 975XG and Enermax 535watt - no issues
DFI P35 LP-UT and Enermax infiniti 650 watt - no issues

All overclock extremely well, although I have heard of a few people with issues .

[Praz]

Although a quality power supply, pairing an Enermax with a DFI board is always a roll of the dice. And more often then not the end-user loses.

[CyberDruid]

Running mine on a PC Power 750 and I still get an occasional Blue Screen after days of steady Folding at 100% load...obviously not quite right despite nice temps and no other signs of instability when stress testing.

[MacKaquE]

hey guys,

I was damn busy this week end so had no time to test and work more on my OC

I did a 10 hours prime last nite ...

3 core were fine for 10 hours

1 core got an error at the sixth hour

Anyone know why 3 core works and 1 don't ?

I'm at 3.5GHZ right now

So I need to give more vcore to the CPU ? If yes how many more I should try first let me show you my setup atm

Thanks for you helps guys !

[Seanie's Show]

@ Brother Esau: I think its Silverstone who Retail the decathlon Series PSU's not enermax.

@ Cyber Druid: I have to ask, if you are still getting BSOD's, make me wonder, how long did you stress test for ?

[Seanie's Show]

hmm thx for the info so far, i`ve got the Decathlon 750w could this be the problem ? Same sqeeling occurs with DFI 680i but not as loud.

I used to have an OCZ ModXStrem 780w before my current one, and either it wasnt enough power for my system or the PSU and this DFI mobo were in love with each other, cos they used to both resonate and sing in unison together , the resonating on the board seemed to come from the memory socket area, however like I said before, since I changed the PSU to a more powerful one, all the singing and whinning has gone, the system just seems generally more stable.

[Brother Esau]

@ Brother Esau: I think its Silverstone who Retail the decathlon Series PSU's not enermax.

@ Cyber Druid: I have to ask, if you are still getting BSOD's, make me wonder, how long did you stress test for ?

Hmmm misread thought I saw Liberty

[CyberDruid]

@ Brother Esau: I think its Silverstone who Retail the decathlon Series PSU's not enermax.

@ Cyber Druid: I have to ask, if you are still getting BSOD's, make me wonder, how long did you stress test for ?

Not long enough

I got WUs to Fold man...Folding@Home at 3600 is an ongoing stress test

But seriously I have been so busy with the mods and the shop I have been neglecting my overclocking duty to this fine board and will get right on it

[emoners]

hmm thx for the info so far, i`ve got the Decathlon 750w could this be the problem ? Same sqeeling occurs with DFI 680i but not as loud.

DA750w user here...Rock solid!

[ic3m4n2005]

That's curious since i thought 750w should be enough.
But one thing is right, the noise seems to come from around the mem area.

@emoners: which Rev. of the psu did you own ? If i remember correctly there are some Rev. which do and some which don't sqeal/whine/whatever


P.S. Stability doesn't seem to be an issue...have to examine this soon.

[emoners]

That's curious since i thought 750w should be enough.
But one thing is right, the noise seems to come from around the mem area.

@emoners: which Rev. of the psu did you own ? If i remember correctly there are some Rev. which do and some which don't sqeal/whine/whatever


P.S. Stability doesn't seem to be an issue...have to examine this soon.

sorry, didn't know there were different revs on this psu. but no squealing here.

[ic3m4n2005]

Revision should be on a sticker on the psu, will examine this when i get home fro m work.
I have to say that i've read this if i remember correctly so this is no proven fact until i could check it on my own psu or someone else says something about it.
If you have access to your psu it would be nice if you could search for the Rev.

[Praz]

So I need to give more vcore to the CPU ? If yes how many more I should try first let me show you my setup atm

You can try more VCORE. Maybe 1.4 volts? Be best to determine if the same core is consistently failing. Prime won't show this. The order of the threads do not match up with the cores. To determine the core that has failed you will need to look in Task Manager.

Same core, most likely VCORE. Switches between same two cores on same die look at CPU GTL for that core. Random failure of all cores, NB voltage or NB GTL.

The above suggestions are not all-inclusive. Other settings/voltages can be at fault but not as likely. And this is assuming the failures are not memory related.

Edit:

On the subject of GTL settings let me know if this would be of use to anyone.

As most of us now know the 110, 110, 110 GTL values displayed as default mean little. The actual values change based on the corresponding VTT and NB voltages. And the GTL value represents a different voltage as the VTT or NB voltage is changed.

Oskar was kind enough to supply us with a partial table of CPU GTL. I had thought about putting together a chart of all the available voltages with the default GTL values. This would fill in the missing values of Oskar's chart.

If anyone has a need for this I can probably get it done over the weekend.

[cuhaos]

I've been using the DA750 with this board and its still squeaks under load. The DA750 is Revision V3.

[ic3m4n2005]

Oh another one, thx for sharing information on this, will add mine in a few hours

[Brother Esau]

You can try more VCORE. Maybe 1.4 volts? Be best to determine if the same core is consistently failing. Prime won't show this. The order of the threads do not match up with the cores. To determine the core that has failed you will need to look in Task Manager.

Same core, most likely VCORE. Switches between same two cores on same die look at CPU GTL for that core. Random failure of all cores, NB voltage or NB GTL.

The above suggestions are not all-inclusive. Other settings/voltages can be at fault but not as likely. And this is assuming the failures are not memory related.

Edit:

On the subject of GTL settings let me know if this would be of use to anyone.

As most of us now know the 110, 110, 110 GTL values displayed as default mean little. The actual values change based on the corresponding VTT and NB voltages. And the GTL value represents a different voltage as the VTT or NB voltage is changed.

Oskar was kind enough to supply us with a partial table of CPU GTL. I had thought about putting together a chart of all the available voltages with the default GTL values. This would fill in the missing values of Oskar's chart.

If anyone has a need for this I can probably get it done over the weekend.

Good Morning , John
Yea Post or pass that Table my way. I know you have been working very hard on this and I would be interested in what you have come up with.

Get my OCZ Flex today and I can put you're findings right to work

[Seanie's Show]

Good Morning , John
Yea Post or pass that Table my way. I know you have been working very hard on this and I would be interested in what you have come up with.

I 2nd that, would love a copy too.

[emoners]

You can try more VCORE. Maybe 1.4 volts? Be best to determine if the same core is consistently failing. Prime won't show this. The order of the threads do not match up with the cores. To determine the core that has failed you will need to look in Task Manager.

Same core, most likely VCORE. Switches between same two cores on same die look at CPU GTL for that core. Random failure of all cores, NB voltage or NB GTL.

The above suggestions are not all-inclusive. Other settings/voltages can be at fault but not as likely. And this is assuming the failures are not memory related.

Edit:

On the subject of GTL settings let me know if this would be of use to anyone.

As most of us now know the 110, 110, 110 GTL values displayed as default mean little. The actual values change based on the corresponding VTT and NB voltages. And the GTL value represents a different voltage as the VTT or NB voltage is changed.

Oskar was kind enough to supply us with a partial table of CPU GTL. I had thought about putting together a chart of all the available voltages with the default GTL values. This would fill in the missing values of Oskar's chart.

If anyone has a need for this I can probably get it done over the weekend.

That would be VERY VERY helpful.

[emoners]

Good Morning , John
Yea Post or pass that Table my way. I know you have been working very hard on this and I would be interested in what you have come up with.

Get my OCZ Flex today and I can put you're findings right to work

would really love to know on the performance of your OCZs vs the Crucials on this...getting some OCZs these weekend

[pump101]

does anyone know wat the 'exit setup shutdown' feature in the bios of this DFI board is for??

thanks in advance for any replies.

[Praz]

does anyone know wat the 'exit setup shutdown' feature in the bios of this DFI board is for??

thanks in advance for any replies.

When the board first boots there are settings for the chipset that need to be configured based on CPU speed, FSB and memory divider. This is why there is a double boot on system startup. When the board is first powered on these three values are loaded, the system reboots, then the chipset is configured with predefined settings based on the combination of these values. As long as the board is not powered down this configuration is retained

Now how this applies to the above Shutdown Option and warm boot. As already noted the chipset settings cannot be applied until the CPU speed, FSB and memory divider are set. If any of these are changed in the BIOS, saved and the system rebooted a double boot needs to take place. DFI has provided a partial workaround for this with the Mode 1 and Mode 2 options.

Mode 1 - Upon rebooting from the BIOS a check is performed against the previously used configuration and the new one. If they are close enough that the currently loaded chipset settings will work the board will boot normally. If there is enough of a discrepancy between the two configurations then the board will double boot in order to initialize the chipset properly.

Mode 2 - Regardless of the changes made to the CPU speed, FSB or memory divider the board will always double boot.

While making changes to the CPU speed, FSB or memory divider it is best to leave this option set to Mode 2. This way the chipset will always be initialized properly. When tweaking other settings Mode 1 works fine.