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

[fireice2]

Thanks. I really need to know this GTL stuff since I don't want to use brute force.

I feel quite uncomfortable feeding my proc high voltages since a lot of people are killing 45nm left and right.

[emoners]

them Wolfies are kinda hard to tame for me... i'm having a hard time in stability

the VTT & GTL are so different from the 65nm

wish Praz would release another table for these 45nm chips like they did with the 65nm ;-)

[MikeMK]

wish Praz would release another table for these 45nm chips like they did with the 65nm ;-)

Yup, I think we would all agree with that! Would be really useful having a meaning full starting point with these chips.

[jimmyz]

Can't say if it is the same for wolfies but my Yorkfield loves it when the vtt matches the cpu volts, lower causes instant crashes at the same speeds if I load all cores.

[karateo]

what max VTT should I use on a E8400 at max 4200?
vcore about 1.35-1.38
VTT I think 1.2 max BUT need to confirm?

When does it start to harm cpu? more than what volts?

[MikeMK]

Can't say if it is the same for wolfies but my Yorkfield loves it when the vtt matches the cpu volts, lower causes instant crashes at the same speeds if I load all cores.

I take it you mean, for example, that if your vcore is at 1.37v, your VTT would be around 1.37v? If so, I may give that a go with my yorkie.

What are your GTL's coming out at? Do they roughly match the GTL scale over at anandtech? or are they way off that, like some of the Wolfdale clocks I've seen?

Some info would be great... lots of ppl with Wolfdales around here, but not so many with yorkfield...

[jimmyz]

I take it you mean, for example, that if your vcore is at 1.37v, your VTT would be around 1.37v? If so, I may give that a go with my yorkie.

What are your GTL's coming out at? Do they roughly match the GTL scale over at anandtech? or are they way off that, like some of the Wolfdale clocks I've seen?

Some info would be great... lots of ppl with Wolfdales around here, but not so many with yorkfield...

exactly, as for the GTL I have it disabled for now. I started tweaking once and stability got worse and i haven't opened it or studied about it since. right now I am on the x-38 board but I have my p-35 running a E6600.

The reward in matching the vtt to the cpu volts is that it allowed me to use lower voltage to acieve stability at the same clocks. so far she can do 5050 mhz 3Dmark stable with 1.5675 set in bios 4.4 occt stable @ 1.3750 cooled with phase of course. even though I use phase I have set a limit for short term benching of 1.6v that can get a screen of 5220 MHz but for daily i keep it under 1.45v since it is so stable I really never got into the gtl stuff yet, this is my first quad and it clocks like a dream.

[MikeMK]

Yup, very nice indeed. Interesting info there. Gonna give it a go myself I think, and see where it gets me.

I think I really need to put some cash together for a phase unit lol! I would really like to be able to hit some of those clocks your getting. Ive been tempted many times before, and chickened out at the last minuite... I have a spare case here... might build myself a cheap practice rig and a phase unit to give it a go first, before I stick it on an extreme edition!

[cool_case]

The reward in matching the vtt to the cpu volts is that it allowed me to use lower voltage to acieve stability at the same clocks. so far she can do 5050 mhz 3Dmark stable with 1.5675 set in bios 4.4 occt stable @ 1.3750 cooled with phase of course. even though I use phase I have set a limit for short term benching of 1.6v that can get a screen of 5220 MHz but for daily i keep it under 1.45v since it is so stable I really never got into the gtl stuff yet, this is my first quad and it clocks like a dream.

Very interesting - thanks! I can't find the VTT setting. Does it go by another name in the BIOS?

[MikeMK]

Well, ive just tried it out, for my attempts at 4.2ghz over the last hr or so. Doesnt work for me. Even went as high a 1.42v, with 1.42VTT and prime failed within about 10mins.

I would be suprised if it needed that kind of jump in voltage, seeing as its 11hrs prime stable @ 4ghz 1.32v. I get the feeling i must be missing something! LOL!

[kberguy1]

Now I am working on tightening memory a bit more, and hopefully reducing vCore and vDIMM at least a notch or two.

For those interested, my GTL settings are:
CPU GTL 0/2= 117
CPU GTL 1/3= 116
NB GTL= 119

I tried to get a stable overclock with 8 x multiplier, but hit a bit of a wall at 465 FSB max. This was with loose memory timings, performance level 7. The memory was working just fine, with the settings I used being stable with Memtest to 1270 MHz.

Along with this 465 MHz FSB "wall" came a need to radically change GTL settings for stability. My original settings shown above had to change to 123-122-124, otherwise the Prime95 test would fail in 1st 10 minutes (and usually the 1st minute or so).

A final note/question: As soon as I tried going above 1.3875 in BIOS, the Motherboard would cycle C1 and off, C1 and off repeatedly. Why is it that I cannot add more voltage? My other CPU (B3 stepping) never had this issue

[=[PULSAR]=]

is the only difference between the LP and UT the transpiper?

[MikeMK]

Yes, as far as im aware they are identical appart from that!

[kberguy1]

In the last week to ten days, I have probably spent 30 hours tuning my Q6600 processors with this motherboard. It struck me that one of the things I was noticing was that as the GTL levels get close to optimum levels, the temperature diffences between the cores tend to be reduced.

To me, this means that some of the people who have been lapping their IHS to reduce/even out temps should be looking at their GTL settings.

Has anyone else been noticing this, or an I seeing a pattern that really isn't there?

[cantankerous]

I personally haven't noticed this no. I usually leave GTL disabled with the buffer strength set to weak, however I know what to set them to to have my system stable at the same clocks and volts and the temps between cores are the same whether I use GTL or not. What you find is interesting if it isn't just happening for you.

[nikolas61]

Super pi 1M 9.71 sec on air

[Delemon]

Hello all,

I recently disassembled my PC to apply new TIM (MX-2) to my CPU, NB, SB, and PWM, and added some heatsinks to my CPL chips. When I reassembled my PC, I was getting stuck in a continuous reboot loop. The system gets as far as POST code 26 before reboot.

I currently have one RAM module installed in the yellow slot closest to the CPU (green slots give me a C1 boot error). One vid card installed, one keyboard, no HD, no CD-ROM. All power connections are plugged in (24-pin, 8-pin, and both floppy connectors). TRUE with a Yate Loon and one smaller 80mm exhause fan are currently plugged into fan headers on the mobo. Running the 12/24 Bios, and I have done a CMOS clear. Cannot get past post code 26 to even get into BIOS.

Any ideas?

[Roger_D25]

I assume everything was working fine before you disassembled the motherboard to re-apply new TIM? Do you feel pretty confident that nothing got broken or damaged during the disassembly or reassembly of the heatsinks? By the way have you figured out what POST code 26 is yet, that is probably a great place to start and will save some testing time? Plus besides just replacing the stock TIM you didn't replace any hardware or do a BIOS change? Interesting, I'm curious how this works out so keep us updated if you would?

[cantankerous]

I would have to say something has gone wrong since mucking with the board as well, as long as you know it was working before hand that is. MX-2 isn't conductive so I wouldn't say a short from the paste is happening, but perhaps some physical damage happened, even with your realizing? Could part of the heatsink have caused a short?

[kberguy1]

I have been posting a few results about my new Q6600 processor. It is running very well for me, but I am stuck at 3780 MHz Prime 95 stable (7 hrs and counting) because I cannot increase my voltage for vCore.

I am currently at 1.374 volts. If I increase to either of the next two points (1.3825/1.3875) I am stable to 3825 MHz, BUT there is no guarantee that my PC will reboot properly. It usually flashes "C1" for a half second or so and then "88". It will continue to do that for 8-10 times, and then it shuts off. About half the time, I can get back in BIOS if I hold down the "home" key; the rest of the time I need to clear CMOS.

I was able to put up to 1.425 v through my old B3 CPU, so I am not understanding what the problem with this new one. Are the G0 stepping Q6600's voltage limited in some way, or do you think my motherboard is failing under the new speeds it is experiencing? Is there something I am overlooking here?

Note: the issue is NOT being caused by excessive temperatures...

[karateo]

i got prime stable at 4158 with following settings:

CPU VID Control: 1.3625
CPU VID Special Add: 102.83
DRAM Voltage Control: 2.19
SB 1.05V Voltage: 1.15
SB Core/CPU PLL Voltage: 1.75
NB Core Voltage: 1.37
CPU VTT Voltage: 1.12
Vcore Droop Control: Enabled
Clockgen Voltage Control: 3.45v
GTL+ Buffers Strength: Strong
Host Slew Rate: Weak
GTL REF Voltage Control: Disable


Do you believe i am overvolting for this kind of overclocking the bold ones?
I have no clue what the special vid is, but I upped it a bit :p

[Delemon]

It is entirely possible that something was damaged, but if there is something damaged, I can't see it readily. I've disassembled again to visually inspect each of the chips and nothing is visually identifiable. I'm going to reapply this stuff and try again.

I've looked into boot codes but all I've found is something about "Enable A20 Line", and I honestly don't know what that entails.

[soundx98]

i got prime stable at 4158 with following settings:

CPU VID Control: 1.3625
CPU VID Special Add: 102.83
DRAM Voltage Control: 2.19
SB 1.05V Voltage: 1.15
SB Core/CPU PLL Voltage: 1.75
NB Core Voltage: 1.37
CPU VTT Voltage: 1.12
Vcore Droop Control: Enabled
Clockgen Voltage Control: 3.45v
GTL+ Buffers Strength: Strong
Host Slew Rate: Weak
GTL REF Voltage Control: Disable


Do you believe i am overvolting for this kind of overclocking the bold ones?
I have no clue what the special vid is, but I upped it a bit :p

CPU VID Special "multiplies" your VID.
So you are at 1.401v (1.3625 x 102.83%)

[karateo]

i read this post
http://www.xtremesystems.org/forums/...&postcount=642
which confused me!

[soundx98]

What George was saying is your settings could also be achieved by using 1.3875v x 0.023% (close example).
The nice thing about the two adjustments are the ability to "fine tune" the VID.
It's also the bad thing as it can affect other settings.