DFI Lanparty UT P45-T2RS Banter starts here!!!

[Jor3lBR]

Q9650 in da house

Only had it 2 hours & I can tell already we are going to get along just fine together.

Yeah you will enjoy the Q9650, by far the best processor I've seen.
My Q9650 needed 1.15vcore for 4005Mhz and 1.10v for 3600Mhz.

Wish you luck on the P45 500x9 is the goal Your problems will start at 9x485 that's when it gets really tuff for the X48 to keep up and stabilize.

Right now on the X48 my 24/7 config 4252Mhz (474x9) using 1.336 on the Digital Multimeter / 266/667 Strap PL 6:

[C-N]

Yeah you will enjoy the Q9650, by far the best processor I've seen.
My Q9650 needed 1.15vcore for 4005Mhz and 1.10v for 3600Mhz.

Wish you luck on the P45 500x9 is the goal Your problems will start at 9x485 that's when it gets really tuff for the X48 to keep up and stabilize.

Right now on the X48 my 24/7 config 4252Mhz (474x9) using 1.336 on the Digital Multimeter / 266/667 Strap PL 6:

http://i35.tinypic.com/2q8ol68.jpg

Thanks bud, I know everyones got one & I know 4GHZ is a breeze but the last quad I had was a Q6600 that did about 3.8GHz so the Q9650 is blowing me away.

I am gonna rap it in for the night its 1/2 one in the morning here (time flies ay) I have her running at 4.3GHz 1.328v loaded right now & she's stable but I'm starting to bully the board a bit with voltage so I'll quit now & start by tweaking the GTL's out tomorrow.

Sure I'll shoot for 500 x 9 but I'll probably aim for about 4.2GHz for daily use. I'll most likely stay on the 333/800 strap/divider & run 4 x 1gb sticks. The P45 boards PL is one behind the P35,X38,X48 etc.

[Jor3lBR]

Thanks bud, I know everyones got one & I know 4GHZ is a breeze but the last quad I had was a Q6600 that did about 3.8GHz so the Q9650 is blowing me away.

I am gonna rap it in for the night its 1/2 one in the morning here (time flies ay) I have her running at 4.3GHz 1.328v loaded right now & she's stable but I'm starting to bully the board a bit with voltage so I'll quit now & start by tweaking the GTL's out tomorrow.

Sure I'll shoot for 500 x 9 but I'll probably aim for about 4.2GHz for daily use. I'll most likely stay on the 333/800 strap/divider & run 4 x 1gb sticks. The P45 boards PL is one behind the P35,X38,X48 etc.

Great! Yeah 4200-4250 is ideal with low volts, for me to run 4300 stable was a big jump on volts those extra Mhz won't be missed. For 485x9 the max PLL I used was 1.8v I still can't put myself on using more then that. Less then 485fsb nothing more then 1.6 was required.

My board likes a little bit less then 67% x Vtt let's trade some feedback to see how it plays on the P45.

Anyway thanks for the info and have a good night.

[C-N]

Great! Yeah 4200-4250 is ideal with low volts, for me to run 4300 stable was a big jump on volts those extra Mhz won't be missed. For 485x9 the max PLL I used was 1.8v I still can't put myself on using more then that. Less then 485fsb nothing more then 1.6 was required.

My board likes a little bit less then 67% x Vtt let's trade some feedback to see how it plays on the P45.

Anyway thanks for the info and have a good night.

I haven't got it nailed completely yet & its only about an hour stable but my chip has roughly run 4GHz 1.216~1.232, 4.1 @ 1.248, 4.2 @ 1.28~1.296 & I'm currently good at 4.3 with 1.328~1.344 loaded according to CPU-Z so its scaling nice.

Haven't touched PLL yet & don't intend to, call me chicken but I've wrecked good chips with PLL & despite what anybody says here on XS about killer voltages IMHO PLL hurts & I'll stay away from it.

My CPU likes 0.67 & so did my NB to a degree but I found I can dramatically reduce the VTT needed by raising NB GTL to 0.69

@ Thule welcome to XS

I would try all GTL's to 0.67 & give it 1.25v vtt then see what happens.

[Thule]

Hey C-N thanks for your reply pal.I got worker number 4 error with prime95 when my cpu vtt 1.356, all GTL's are 0,67 and cpu voltage 1.36.I tried everything but nothing is changed.

[DJSUB]

I haven't got it nailed completely yet & its only about an hour stable but my chip has roughly run 4GHz 1.216~1.232, 4.1 @ 1.248, 4.2 @ 1.28~1.296 & I'm currently good at 4.3 with 1.328~1.344 loaded according to CPU-Z so its scaling nice.

Haven't touched PLL yet & don't intend to, call me chicken but I've wrecked good chips with PLL & despite what anybody says here on XS about killer voltages IMHO PLL hurts & I'll stay away from it.

My CPU likes 0.67 & so did my NB to a degree but I found I can dramatically reduce the VTT needed by raising NB GTL to 0.69

@ Thule welcome to XS

I would try all GTL's to 0.67 & give it 1.25v vtt then see what happens.

we're the same. i also wrecked 2 good chip e8400 by upping the pll voltage to 1.75v for hitting 600fsb on P35. the chip died in less than 5mins without doing some benchmarks or anything. i don't know if it's connected with a higher vtt. but i think P45 is a different story. i saw some benchmark using high pll voltage & i think they never had that problem. but i'm still afraid getting the pll voltage up even on P45 because of that bad experience

[C-N]

Hey C-N thanks for your reply pal.I got worker number 4 error with prime95 when my cpu vtt 1.356, all GTL's are 0,67 and cpu voltage 1.36.I tried everything but nothing is changed.

Where has it crashed, what fft & does it always crash there?

I also find setting the GTL Buffer Strength to weak helps my chip.

we're the same. i also wrecked 2 good chip e8400 by upping the pll voltage to 1.75v for hitting 600fsb on P35. the chip died in less than 5mins without doing some benchmarks or anything. i don't know if it's connected with a higher vtt. but i think P45 is a different story. i saw some benchmark using high pll voltage & i think they never had that problem. but i'm still afraid getting the pll voltage up even on P45 because of that bad experience

I have seen plenty of people using high PLL & getting away with it but I'm not taking any chances. P45 may be different & I have very briefly tried upping the PLL on this UT P45 with my quad to see if it could stabilise 500 x 9 which it didn't so luckily for me I should have no reason to run PLL at anything higher than stock.

For anyone interested the PLL & VTT measure points are measured on the chip marked SK12 directly behind the yellow SATA ports. The D6 side of the chip is PLL & the other VTT.

[C-N]

I'm still not fully comfortable with my quad oc atm (still learning fast) but thought I would post up where I'm at right now just to show you quads do work on this board.

Q9650 @ 4.2GHZ 2 X 1GB @ DDR1167 / 266/667 Strap/Divider.

[Thule]

hi C-N .here is the picture.

[C-N]

hi C-N .here is the picture.

Ouch thats instant fail, you have a long way to go.

I would back the fsb down, find a stable point & work up from there.

[Jor3lBR]

I'm still not fully comfortable with my quad oc atm (still learning fast) but thought I would post up where I'm at right now just to show you quads do work on this board.

Q9650 @ 4.2GHZ 2 X 1GB @ DDR1167 / 266/667 Strap/Divider.

Nice clock but I was just hopping the P45 had a better PWM design able to support Quads power amount. Looks like they still kept the fragile design used on the X38 and X48 series (for economic reasons), this is really a pity as they work wonders for Duo's but can't scale with Quads.

Those Q9650 are really comfortable along the 500fsbx9 line with aircooling so not to use this amount of extra power is just a waist.

But in any case if you're generally happy with 460-480fsb the DFI's are solid stable with low amount of volts but they will just fail above that.

Thanks for the effort CN!

[C-N]

@ Jor3lBR,

The boards not a slouch with a quad in any way but it still cant out do any of its predecessors P35, X38, X48 etc.

I see other boards making 500fsb with a quad but I don't see the masses holding 500 stable.

I don't think the issue is the PWM there may or may not be an issue with the PWM that I'm not aware of but for me the issue of quads not scaling up like for arguments sake the UD3 has more to do with the "CPU Clock Amplitude". My board/quad needs this to be increased from 800 to 900 to even boot over 475fsb & it needs it upped again to 1000mv to go past 500fsb.

I don't know if its a hardware limitation or even if 1000mv is insane & likely to kill a chip or not but I sure would like to pump a bit more to see what happens.

Oh... & 460 ~ 480fsb is plenty for daily use but if 500+ is available on the UD3 then it'll be a UD3 I'm ordering simple as.
.
.
.

@ namegt looking good.
.
.
.