** Official ** Gigabyte UD3 P45 Series -- EP45-UD3 / EP45-UD3R / EP45-UD3P

[Lsdmeasap]

Glad it could help a bit, I think it should be added to page one for all to see it for sure!

I mod the Gigabyte Forum at TweakTown, so if you have any questions feel free to shoot them at me

My name there is Lsdmeasap, I was going to use that here but thought it was the reason my activation/approval did not go thru so I registered again with a different name

[MacClipper]

Got hold of the rev 1.1 of the UD3P for some testing today and it's a world of difference with regards voltage stability. Thought you guys might be interested.
Rev 1.0 and 1.1 compared,
http://forums.vr-zone.com/showpost.p...9&postcount=42
Rev 1.0 - earler findings Oct '08,
http://forums.vr-zone.com/showpost.p...23&postcount=4

Might have been a bad batch of rev 1.0 mobo we had locally a few months back though.

[PolRoger]

I have been running this O.C. for a little while now... I like it because of the low volts which makes it nice for 24/7. Loadline enabled, 333 latch, 2.40B divider, 2 sticks of Kingston HyperX and a TRUE Black for cooling. For the stability test I ran Prime custom FFT at 1024K to 4096K.

[pangingIII]

@mark0326,
personally, i think, from now on, if people post the ET6 window along with their overclocking achievements then it will be more than enough information.

==========================


this would probably be my first choice for an every day booting/setting for this Q9300 on a UD3R board.
CPU with True and 2 12mm mid rpm fans, the overall system temperature was seems to be fine.

Q9300 480x7.5 5:6 DDR2-1152.
i added more vcc,vtt and mch core volts to my last settings (in post #1111), in that one i ran superpi 32mb's tested only.

this time i run Prime95 blend (still 2hr test only)...




==========================

btw, MacClipper, thanks for the info, that's good information and good to know.

[bluestang]

a573573 and PangingIII

Thanks for your detailed explainations and postings. Very valuable info. I think I'm starting to understand now how ET6 works compared to BIOS. ET6 show the %, not the volts like BIOS. For example, using PangingIII ET6 window above, his MCH Ref is .76 and vtt is 1.34. So in BIOS it would show what as far as volts then?

[Lsdmeasap]

Well The BIOS shows % as well, but in volt format. .63% to be exact of Vtt for MCH ref unless you deviate from that percentage

Please do notice that at 1.20 Vtt stock, MCH Ref and CPU ref for example are .76 same as you would see in ET, which is .63% of Vtt. You have to keep the default % values in mind

As for your question about the above, with .76 showing in the ET Screen and 1.34Vtt, the BIOS MCH Ref could be set on Auto or it would be manually set at .8442

[bluestang]

So does auto work well in BIOS if I have to up my Vtt volts and MCH Core volts to be stable? Or is manual the way to go?

[pangingIII]

bluestang,



ok, i don't remember the value in bios and never want to remember it, lol, i'll have to reboot the system and to look in bios in order to answer your question.
so i hope you won't mind taking this example?


so here's a example...

once im in bios, i set CPU term. (VTT) to 1.20
DRAM volts to 1.8

then go to these...
... CPU PLL and set its value to 1.50
... CPU ref. 0.83
... CPU core to 1.35

... DRAM term. 0.81
... Ch A ref. 0.81
... Ch B ref. 0.81

... MCH core 1.34
... MCH/DRAM ref. 0.81
... MCH ref. 0.76
... ICH I/O 1.59
... ICH core 1.1

then, after that i go back to the CPU term. and change the 1.20 setting to 1.34
and then change the DRAM volts from 1.8 to 1.9
and once you boot into Windows that will be the voltage value you see in the ET6.

i normally don't adjust voltage setting in ET6, i only use it sometimes but then i want to reboot the system with a new setting in BIOS before i actually test the system.
so if i may i would suggest you do the same.

[mark0326]

@mark0326,
personally, i think, from now on, if people post the ET6 window along with their overclocking achievements then it will be more than enough information.

Maybe I should ask for help with my overclock then rather than trying to look for the answers by myself and not understanding all of what I read. Guess pride is getting in the way, lol.

[Lsdmeasap]

So does auto work well in BIOS if I have to up my Vtt volts and MCH Core volts to be stable? Or is manual the way to go?

Ya, you can leave Ref Voltages on Auto if you like for sure. It may or may not be stable, but you can leave them. Don't leave Vtt or MCH on Auto though, you would want to set those manually.

If you Leave Ref voltages at Auto they will always be this amount in ET no matter what you set Vtt to, but the actual voltage used will be the respective .xx% amount of Vtt >>>

CPU Reference Voltage (.63% of Default Vtt/1.20) .76 ET
CPU Reference2 Voltage (.67% of Default Vtt/1.20) .80 ET
MCH Reference Voltage (.63% of Default Vtt/1.20) .76 ET

[pangingIII]

Ya, you can leave Ref Voltages on Auto if you like for sure. It may or may not be stable, but you can leave them. Don't leave Vtt or MCH on Auto though, you would want to set those manually.

If you Leave Ref voltages at Auto they will always be this amount in ET no matter what you set Vtt to, but the actual voltage used will be the respective .xx% amount of Vtt >>>

CPU Reference Voltage (.63% of Default Vtt/1.20) .76 ET
CPU Reference2 Voltage (.67% of Default Vtt/1.20) .80 ET
MCH Reference Voltage (.63% of Default Vtt/1.20) .76 ET

try 600fsb and left MCH ref. at auto.
MCH ref. not always remain at .76v value. the value will be changed on certain CPU FSB as far as i can see.

[Lsdmeasap]

Ahh, I was not aware of that as I have no Ln2 to run 600 FSB!

Can you document these changes and at what points so I may add them to my guide/notes I post about this

[BuGjAE]

Voltage Instabilty in UD3P REV 1.0 OMG !!!! ,This is cause to get error while run Orthos for sometime. Is that right? I got freezing after run orthos over 2-3 Hrs.
Now I'm using [520x8.5] with 1.360 Vcore in CPU-z .I'm never noticing the Vcore graph before. Do you guys think this causing unstable?

enable LLC
CPU Enhanced Halt (C1E)...................: [Disabled]
C2/C2E State Support......................: [Disabled]
x C4/C4E State Support....................: [Disabled]
CPU Thermal Monitor 2(TM2) ...............: [Enable]
CPU EIST Function.........................: [Disabled]
Virtualization Technology.................: [Disabled]

[pangingIII]

Ahh, I was not aware of that as I have no Ln2 to run 600 FSB!

Can you document these changes and at what points so I may add them to my guide/notes I post about this

as far as i can remember the MCH ref. value of 0.76 will be changed to 0.74 maybe at above 400-450fsb, also, the value even change from 0.76 to 0.74 when you use ET6's quick boost to overclock an E8400 to 450fsb.
you can try that if you always set it on auto, of course you will need to check the value in ET6, this due all you will see in bios is Auto (and no value).
however, after the first couple days i recieved a UD3R board i never set this bios parameter on auto, so i can be wrong and it's better that you check it yourself.
i've been doing alot of overclock testing on a UD3R board and E8400 before, but not a lot of reading, so may be i will read yours when you are done with it.

[Lsdmeasap]

Well it would have to be above 450 as that is what I run normally. I have ran 533+ but never looked, I'll have to look into this tonight when I get some time! Thanks for sure for pointing that out as I had no idea

I cant do 600 FSB but I can for sure do 533-560

I have a P45-DS4P and a E3110, so we have the same BIOS basically and CPU. I think the only difference in BIOS between our boards is I have CPU ref 2, the rest is the same.

I will check out MCH Ref in Auto at High FSB, you may very well be right

[pangingIII]

i don't know about your board model as i've never tried one, but UD3R's and 600fsb was one of the very first i tried on the board once i have it, so i think i remember that one, i first tried with all volts and ref. setting on auto, then booted into windows and took a look at ET6's value, the MCH ref. was set to 0.65v.

[Lsdmeasap]

Ahh, well I have never tried it like that, it will probably do 600 FSB then at Auto settings as I do know Vtt and Pll are into the 1.8 range at 450 FSB if left in Auto.

I'll give it a few shots in auto and a few manually set and see how she does. My Xeon is a volt hog though, so I may not want to push the vcore needed, I dunno

[ghost_recon88]

I can't understand this board It's constantly running my Q9550 quite a bit hotter then my TPower I45 did. On this board, at stock clocks (2.83GHz) with only I'm hitting 48c. I know on the TPower I had the CPU GTL ref at .63x, and on this board it defaults at .76x. Could that be the difference? I'm using the F4 BIOS. I've reapplied thermalpaste and reseated my TRUE several times.

TPower I45


EP45-UD3P

[Homeless2]

Got hold of the rev 1.1 of the UD3P for some testing today and it's a world of difference with regards voltage stability. Thought you guys might be interested.
Rev 1.0 and 1.1 compared,
http://forums.vr-zone.com/showpost.p...9&postcount=42
Rev 1.0 - earler findings Oct '08,
http://forums.vr-zone.com/showpost.p...23&postcount=4

Might have been a bad batch of rev 1.0 mobo we had locally a few months back though.

I think all the rev 1.0 mobos are like that. Are there any physical differences between the revisions?

[Homeless2]

This is what my occt vcore looks like after the 7b bios. I'm curious as to which bios comes stock on the 1.1 revision

[MacClipper]

I can't understand this board It's constantly running my Q9550 quite a bit hotter then my TPower I45 did. On this board, at stock clocks (2.83GHz) with only I'm hitting 48c.

Once the Vcore is manually set in the BIOS, Gigabyte keeps the Vcore at the level set even with EIST/C1E activated unlike Biostar/DFI (using CPU VID Add) which drops the Vcore when idle resulting in lower idle temps. Not sure if this is what you are experiencing.

I'm curious as to which bios comes stock on the 1.1 revision

F6

Are the other volt lines, viz. 3.3/5v, stable on your mobo? This question is also directed at BuGjAE too.

[Homeless2]

Using a corsair TX750w if that makes a difference. Any bios other than 7b doesn't have normal ripple in occt tests for me (haven't tried F6 though)

[MacClipper]

Looks good so it seems that there are perhaps both good and bad rev 1.0 mobos out there, wonder if it is fixed for all rev 1.1 mobos. Issue is unlikely to be PSU related cos I was using a proven Corsair HX620 unit which did not have such problems with other mobos tested. Now I am just using a lower end AcBel E2 470 on the rev 1.1 for testing yet the lines are all stable.

[bluestang]

Corsair Dominators 2x2GB PC2-8500 (2.1Volts, version 1.1) should be able to run 1:1 with a FSB of 533 right? I'm having difficulty with settings for 8x533.

[BuGjAE]

@bluestang, sure bro.I think my GEIL 1066 2x2GB can do 1:1 FSB 533 also but I still can't get it stable. even try to increse VTT or MCH core is not affective.