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

[clo007]

I am about to give up on this board. This is my second one (UD3R). The first one came DOA. This second one ran fine till yesterday. I tried to boot up and nothing happened ie. monitors read no signal. I clear battery for 12hrs, as I was at work and such. Put things back together this morning and still nothing. One point yesterday with the issues started, it would not boot. I entered into bios and set to fail-safe settings. Saved and exited. It booted up, went into window, then 30 secs later the computer locked up. Tried to clear cmos/battery again and nothing. Back to today. So I powered on and computer set there cycling. It never went to the bios screen. I thought maybe I was having hdd failure b/c i can't hear them spooling up at all, but I can't even get to the bios without the hdd's installed.

Is my second board failing or what?

Specs of set-up in sig. Thanks guys.

[nullface]

What DDR2 do work nicly on UD3P?

[Ghostleader]

Nice results Panging III

Did you have to spend much time to get those to run at that speed?

Seems like you run those in the orange slots, any special reason for that or is the orange better then the yellow at high speed?

Are you stable with LinX and/or Prime blend with those settings?

[pangingIII]

Nice results Panging III

Did you have to spend much time to get those to run at that speed?

Seems like you run those in the orange slots, any special reason for that or is the orange better then the yellow at high speed?

Are you stable with LinX and/or Prime blend with those settings?

it's a very new ram kit on this board i haven't tested HCI memtest, prime's and Linx on it yet.
however, this ram kit passed a DOS memtesting program (when it's running DDR2-1200) at 1.98v vdimm on all 333 strap (375fsb mem multi 3.2x and 500fsb mem multi 2.4x), 266 strap (400fsb mem multi 3.0x), and 400 strap (450fsb mem multi 2.66x), and i just added a little bit more volts to the bios's vdimm settings and use vMCH at 1.30v (thinking about lowering this value down to 1.26v later on) and booted into Windows and run the tests.
i have 4 other 4GB (2x2GB) kits that won't do these. as of now, it seems to me that it should be possible to stable this ram kit for 500fsb DDR2-1200, and i hope it can make it at least at the 500fsb setting.

about the DIMM slots, the ram is too fat to fit in yellow slots, it hit the board stock sink (heatsink cover).
i actually like to use yellow slots, i find that it works better for my ram overcloking sometimes, so i'll probably also try using yellow slots later on.

[njl]

I am about to give up on this board. This is my second one (UD3R). The first one came DOA. This second one ran fine till yesterday. I tried to boot up and nothing happened ie. monitors read no signal. I clear battery for 12hrs, as I was at work and such. Put things back together this morning and still nothing. One point yesterday with the issues started, it would not boot. I entered into bios and set to fail-safe settings. Saved and exited. It booted up, went into window, then 30 secs later the computer locked up. Tried to clear cmos/battery again and nothing. Back to today. So I powered on and computer set there cycling. It never went to the bios screen. I thought maybe I was having hdd failure b/c i can't hear them spooling up at all, but I can't even get to the bios without the hdd's installed.

Is my second board failing or what?

Specs of set-up in sig. Thanks guys.

Not a definate answer, but are you *absolutely* sure that your OC is 112% stable?

I ask because I'm in the process of OC'ing my own Q9650 - although on a DQ6 mobo, and as you probably know, small changes can make or break the stability.

If your board is cycling at boottime, it's usually an unstable setting.

Could also be heat related - do you have proper cooling/airflow?

Another thing you could try is to remove the mobo from your cabinet and lay it all out on a cardboard box or similar and fire it up.

I had great difficulty getting my rig stable when inside my cabinet after exchanging my CPU-cooler and the reason was most likely some sort of short between the mobo and cabinet. The solution was to remove the setup from the cabinet - no problems with stability.

Have actually sold the cabinet now as it was too small anyway and now it's running, dressed in its birthday suit

Hope you find the answer!

[Jor3lBR]

I posted 4350mhz a few pages back with all my settings which could give you a good start. 4400mhz is all I can call/claim I have got truly stable although 4450~4500 is perfectly stable for everything other than prime/orthos/IBT etc.

I have/did notice something divider/strap dependent I think it was performance level getting set to tight or something like that which stumped me for a minute but other than that I've easily pumped 1.75~1.8v through my NB no problem.

Right now I can get my CPU to run a loop of prime small & then fail large or the other way around depending on how I set up the board for 500 x 9. I am real busy in Jan 09 so wont really get to play with the UD3P until the end of the month & then I think I'll pop the board under my SS & see if that helps.

All in all though the UD3P is an awesome board.

Thanks mate, I got it on page 35.

So 'Normal' settings worked better for you then 'Auto' or 'Manual'?

What does this board recognize as Normal?

Sorry for the questions I had to give up my second board and get the first one back as it turns out the MCH core thing was just a strap related problem so I'm waiting again my old board.

[clo007]

Not a definate answer, but are you *absolutely* sure that your OC is 112% stable?

I ask because I'm in the process of OC'ing my own Q9650 - although on a DQ6 mobo, and as you probably know, small changes can make or break the stability.

If your board is cycling at boottime, it's usually an unstable setting.

Could also be heat related - do you have proper cooling/airflow?

Another thing you could try is to remove the mobo from your cabinet and lay it all out on a cardboard box or similar and fire it up.

I had great difficulty getting my rig stable when inside my cabinet after exchanging my CPU-cooler and the reason was most likely some sort of short between the mobo and cabinet. The solution was to remove the setup from the cabinet - no problems with stability.

Have actually sold the cabinet now as it was too small anyway and now it's running, dressed in its birthday suit

Hope you find the answer!

It has been rock solid at 3.6ghz for 2 weeks since I got the new mobo in. Grounding is not an issue as I put electric tape underneath the mobo at the connection points. It does not even boot into bios now.

[njl]

It has been rock solid at 3.6ghz for 2 weeks since I got the new mobo in. Grounding is not an issue as I put electric tape underneath the mobo at the connection points. It does not even boot into bios now.

Have you tried booting with a *completely* minimal setup (no fd, dvd etc.)?

Could it be a faulty PSU?

You can short the green wire with a ground wire (ATX 24-pin plug) and fire up the PSU. Then you can measure the voltages on the 3.3, 5.0 and 12.0V rails and check if anything is out of the ordinary.

[ezaleina]

Huh..finally can run 32MB super PI.,it took me couple of days to set & run.,set & run..everything else pass.,blend.,linpack.,and OCCT..

[pangingIII]

Ghostleader,

just an updated on my previous post...

changed the OS system from 32-bit Vista to XP x64 so later on i can test more ram in windows

also, did some changes in voltage and ref. settinga at BIOS.
then ran a double 32M SuperPi test and the system is now just started to run a prime95 blend test.

[ezaleina]

@Panging III
can i know what is your Vdrop & Vdroop bro.,is Loadline Calibration enabled?

[dccmadams]

Hello, I just this mobo cpu setup last night. It is intel burn test stable for at least 30 passes. I didnt try more yet. Seems to be a nice mobo.

[Lsdmeasap]

.978 is very high ref value, around 81.5% I wounder if that is right.

Should´t not be something like 1,20 x 68% = 0,816 ????


Hope that fatguy1992 respond and can give use an answer to this.

Ya, it sure is! But he asked how the user above him got 81.5 in his Easytune Image (Apparently that is what he was using ATM)

So that is why I told him how to get 81.5 shown in Easytune, by setting that % in the BIOS.

[C-N]

Thanks mate, I got it on page 35.

So 'Normal' settings worked better for you then 'Auto' or 'Manual'?

What does this board recognize as Normal?

Sorry for the questions I had to give up my second board and get the first one back as it turns out the MCH core thing was just a strap related problem so I'm waiting again my old board.

You're welcome, I just set Normal rather than Auto as Auto sets un-necessary high voltages for a few things. I have most of the board set manually now but to be honest it has hardly made any real noticeable difference so far so probably not worth the time.

[clo007]

Have you tried booting with a *completely* minimal setup (no fd, dvd etc.)?

Could it be a faulty PSU?

You can short the green wire with a ground wire (ATX 24-pin plug) and fire up the PSU. Then you can measure the voltages on the 3.3, 5.0 and 12.0V rails and check if anything is out of the ordinary.

Well, I got it up again. Yanked out one stick of my HZs and it posted. Loaded fail-safe settings and booted.

So my question is, what is the issue? I usually have my black pis installed 2x2gig sticks. The first issue occurred and I put in my 2x1gig sticks of HZs. Someone suggested pulling a stick, and it boots now.

So are my ram slots going bad?

[fatguy1992]

Well, I got it up again. Yanked out one stick of my HZs and it posted. Loaded fail-safe settings and booted.

So my question is, what is the issue? I usually have my black pis installed 2x2gig sticks. The first issue occurred and I put in my 2x1gig sticks of HZs. Someone suggested pulling a stick, and it boots now.

So are my ram slots going bad?

Maybe ur rams going bad .

[clo007]

So my set of black pi's and my set of HZs are going bad? That I would find hard to believe.

[mark0326]

C-N, would you recommend BIOS f4 over anything released since for Q9650?
Thanks.

[fatguy1992]

So my set of black pi's and my set of HZs are going bad? That I would find hard to believe.

It was only a thought . I'v never heard of a ram slot dieing, but i guess it could happen .

Edit: When u took a stick out and it worked, did u try putting the one that worked into the slot you think could be dead?

[clo007]

I just tried both sticks of HZ in slots 1 and 3. No boot. Pulled stick from slot 3 and put into slot 1. Booted.
I am going to try slots 2 and 4 now and see what happens.



Edit: So slots 2 and 4 no boot.

I pulled all sticks out and am only running one stick of my Black Pis in slot 1 and it boots. So the issue is, any type of ram will work if it is only one stick. If I use 2 sticks, it won't post to the bios.
Any idea what would cause this?


Bios settings are all stock right now, nothing done to them.

[C-N]

C-N, would you recommend BIOS f4 over anything released since for Q9650?
Thanks.

I tried them all & couldn't find any more fsb so figured as I had spent time with F4 I would just lay a stable clock down with it & work on from there.

[Xvys]

So my set of black pi's and my set of HZs are going bad? That I would find hard to believe.

I'm having the same problems with mine. My UD3R often does this endless non-boot routine with it starting up and stopping. This seems to happen when you try an overclock it doesn't like. Perhaps this is a memory issue...trying to boot with advanced ram timings, but the bios loads at 1.8v? I thought it was my ram, and I did get a mis-matched set to work for awhile.

Don't matter now. The bios is garbage and the board is not refined. I've pulled this UD3 board out and am selling it. I installed a Asus P5k-Premium using all the same parts and have a higher overclock at a lower voltage that the Gigabyte-me. I'm at 4320 MHz @ 1.34v with my Asus and it boots every time, even with a radical o/c...but I guess I should be posting this in another thread from now on.

[clo007]

That's good to hear I am not the only one with this issue.

[pangingIII]

@clo007.

bad board, bad dimm slots or bad ram kit or these 2 specific ram kits have an issue with default bios. can't tell you exactly without trying the hardwares.
but, try 1 stick ram in first dimm slot, clear cmos, and startup the system, if things seems normal during the first startup, then just left bios at default settings and test this ram stick with memtest program in DOS,
whatever results come out of the mem testing. then replace the stick of ram with another stick, clear cmos, start the system and run a memtest on this stck. if both sticks don't fail memtest? the ram kit is most likely ok.

then try this, with the last stick of ram in first dimm slot and you're already sure that it passes the prior memtest, reflash your bios (in DOS mode if you can) with the same bios version or any release bios version's that you like. after the flashing is completed, power off the system, install the 2 sticks ram in the 2 yellow dimm slots, clear cmos, and startup the system. then please tell me what's the results that you get from and after doing all these. apologies for this long post, i don't know how to make it shorter.

[Ghostleader]

it's a very new ram kit on this board i haven't tested HCI memtest, prime's and Linx on it yet.
however, this ram kit passed a DOS memtesting program (when it's running DDR2-1200) at 1.98v vdimm on all 333 strap (375fsb mem multi 3.2x and 500fsb mem multi 2.4x), 266 strap (400fsb mem multi 3.0x), and 400 strap (450fsb mem multi 2.66x), and i just added a little bit more volts to the bios's vdimm settings and use vMCH at 1.30v (thinking about lowering this value down to 1.26v later on) and booted into Windows and run the tests.
i have 4 other 4GB (2x2GB) kits that won't do these. as of now, it seems to me that it should be possible to stable this ram kit for 500fsb DDR2-1200, and i hope it can make it at least at the 500fsb setting.

about the DIMM slots, the ram is too fat to fit in yellow slots, it hit the board stock sink (heatsink cover).
i actually like to use yellow slots, i find that it works better for my ram overcloking sometimes, so i'll probably also try using yellow slots later on.

Sound nice, I thinking of get some 2x2 PC6900 myself, either OCZ or G Skills Pi (would like to have them in black, hope they release some).

Huh..finally can run 32MB super PI.,it took me couple of days to set & run.,set & run..everything else pass.,blend.,linpack.,and OCCT..

Nice OC

I know, I have struggled some also to OC but I start to learn how the board works, you should get it stable with LinX and Prime blend also, should´t be a problem.

Ghostleader,

just an updated on my previous post...

changed the OS system from 32-bit Vista to XP x64 so later on i can test more ram in windows

also, did some changes in voltage and ref. settinga at BIOS.
then ran a double 32M SuperPi test and the system is now just started to run a prime95 blend test.

How did that Prime session went? Any LinX run yet?