**Official DFI LanParty UT X48-T2R(and LT) Discussion

[AriciU]

I've done it and was very surprised when i tried to mount my TRU in the way i like it. I read that it doesn't fit unless you chop off 1cm from the NB heatsink. Guess that's correct only for the TRUE with the 6 heatpipes.

Re-did the SB and PWM with AS5 instead of that stock paste they used and got a 5-7C drop under load. Made it smell nice too with some of Jean Paul's stuff. Happy camper here.











Idle



Load

[SuporterPoli]

Nice results AriciU.
Does anyone have a GTLREF table for this board with 45nm quads?

[Sauron_Vie]

does anyone know the hole spacing of the NB/SB/PWM´s ?

tia

[Dimek]

Nice results AriciU.
Does anyone have a GTLREF table for this board with 45nm quads?

Check this out
http://www.anandtech.com/mb/showdoc.aspx?i=3294&p=14

[axis]

I've done it and was very surprised when i tried to mount my TRU in the way i like it. I read that it doesn't fit unless you chop off 1cm from the NB heatsink. Guess that's correct only for the TRUE with the 6 heatpipes.

The TRUE is the one with 6 heatpipes. The other is just a Thermalright Ultra 120. The "E" is the extreme version.

Nice results AriciU.
Does anyone have a GTLREF table for this board with 45nm quads?

Look at the first post of this thread. There is a link to Clunks GTL references. A necessary read for anyone with this board.

Axis

[AriciU]

Yeah, i know about the heatpipe difference, just didn't expect my TRU to fit at all.

Any idea why there are difference's in comparing Clunk's GTL settings with those from Anandtech?

[Vega]

Nice results AriciU!

Just a question. What smartguardian do u use becuase my allways show atleast 10 degrees more than Realtemp etc? Have u gays change anything in config file or what?

[IMG][/IMG]

[AriciU]

Nice results AriciU!

Just a question. What smartguardian do u use becuase my allways show atleast 10 degrees more than Realtemp etc? Have u gays change anything in config file or what?

I'm using the one that came on the CD. Same one you're using probably. I can't compare idle temps to real temp / coretemp because my cores are "stuck" on 52/52 (real temp), 62/62 (core temp). The load temps that Smartguardian shows are on par with those of Coretemp (when i pass 63C that is), so 10C more then Real temp.

[Vega]

Oki, then it is ok I quess!

One of my sensors was also stuck at 45 degrees but now it´s ok again!

[AriciU]

Oki, then it is ok I quess!

One of my sensors was also stuck at 45 degrees but now it´s ok again!

I have one of the first batches. They are stuck on it untill it dies

[Vega]

I dont manage to mount my cpu block 100 % correct so my temp is way to high. Yours are better with air! Have you seen if some one else have had a skew cpu socket on this board? I start to think my are.

[IMG][/IMG]

[Carsten]

Does anyone know why there is a newer (Beta) bios version, for the DFI UT X48 version, than for the DFI LT X48 version ? Can you still use the newest version on the LT version ?

kindly
Carsten

[zsamz_]

Does anyone know why there is a newer (Beta) bios version, for the DFI UT X48 version, than for the DFI LT X48 version ? Can you still use the newest version on the LT version ?

kindly
Carsten

beta bios site
i'd suggest the ut x48 bios
http://www.dfi.com.tw/Support/Downlo...FLAG=B&SITE=US

[Sauron_Vie]

does anyone know the hole spacing of the NB/SB/PWM´s ?

tia

nobody ?

[Loony]

I was wondering how the UT X48 T3R was at clocking e8400s and DDR3? I'll be watercooling the e8400.

Also, which DDR3 kits are clocking the best on here? Preferably 2x2gb kits.

[SuporterPoli]

Look at the first post of this thread. There is a link to Clunks GTL references. A necessary read for anyone with this board.

Axis

Yeah but those are for the 65nm . Not quite the same thing i can tell you that.

[zanzabar]

Yeah, i know about the heatpipe difference, just didn't expect my TRU to fit at all.

Any idea why there are difference's in comparing Clunk's GTL settings with those from Anandtech?

i dont like clunks its using a really high clock, but i have found that teh lower clocks work better, i use 68% for my dial core, 69.5% for 65nm quads and 62% for 45nm quads and 61 for 45nm duals

the differences are that the gtl is the clock of how long each pwm works then shuts off, the more power draw the more clocks that u need, and u can use a few sets of clocks and clunks are really high but stable and my recommendations are in the middle ground of not being on the low end but not on the high end

like i said u can use a few bass settings its like 22% 44% or 66% or 88% (anything in 22% increments) its just finding what works (im not advising to use those unless u want to toil with them)

[AriciU]

Damn

Tried to bench at 4500mhz, got into windows and then it shut off on me.

It then got into a start, reset, start, reset loop showing post code 88. I tried to clear cmos by jumper, didn't work.. Ended in same loop

I removed battery, now it starts and is stuck on a c1 error. Troed each stick of ram. Still c1 error.

Cpu voltage was 1.55v. I burned last e8500 at 1.65v. These early ones are so sensitive it's not even funny

Anything else i can try other then removing battery again or do u think cpu is dead?

Sorry for spelling, typing from phone

[AriciU]

Yup. Second e8500 dead I'm afraid to oc on dfi now. Does 1.55v bios give something crazy like 1.65-1.7v or Wtf?

[KingOfToasty]

Yup. Second e8500 dead I'm afraid to oc on dfi now. Does 1.55v bios give something crazy like 1.65-1.7v or Wtf?

Did you happen to change the CPU V Special ADD in the voltage settings? I think if you leave it on auto it can add a lot of extra voltage to your VCore. You might try setting it to its lowest setting next time if you haven't already tried that before.

[zanzabar]

Yup. Second e8500 dead I'm afraid to oc on dfi now. Does 1.55v bios give something crazy like 1.65-1.7v or Wtf?

1.55 is death for a 45nm, u shouldent go over 1.45-1.47, and 1.4V VTT

there is no vdroop on this board so u cant use the voltages that the gigabyte users have, that voltage under load is just instant death


also to do a full reset u have to remove and reattach teh atx cables and cycle the psu power affter

[AriciU]

i tested on an asus p5ke on 1.57v real and it had no problems for about 4 hours of benching. I didn't expect it to go this fast at 1.55v to be honest.

[Praz]

the differences are that the gtl is the clock of how long each pwm works then shuts off, the more power draw the more clocks that u need, and u can use a few sets of clocks and clunks are really high but stable and my recommendations are in the middle ground of not being on the low end but not on the high end

GTLREF is the crossover point of a signaling circuit. Really has nothing to do with any power delivery circuit at all.

[zfactor]

praz is there anywhere to learn all that stuff what each really means? id really like to know and im interested in knowing the true meaning of what things like that really are..any good links you know of to explain?

[Praz]

praz is there anywhere to learn all that stuff what each really means? id really like to know and im interested in knowing the true meaning of what things like that really are..any good links you know of to explain?

Although an older document describing Fairchild's implementation of GTL, this pretty much explains the basic concept.

GTLP: An Interface Technology for Bus and Backplane Applications