Gigabyte P67 UD7 "Socket Burn"?

[Kal-EL]

Finally decided to give this chip a rest and move onto the next chip. This is what I found on the bottom of the chip. The corresponding area of the socket also showed slight mirror signs of burn. Does anyone know what circuit duties this area of cpu pads are responsible for?

The chip still does x54 fine and stable but needs cold to do x55.

Max Voltages used:
vcore 1.67
vdimm 1.72
vtt 1.22
level 2 LLC

Max clocks 5.5ghz run

F2 bios

Hi-Res 1 (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3200.JPG)
Hi-Res 2 (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3202.JPG)
Hi-Res 3 (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3203.JPG)
Hi-Res 4 (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3204.JPG)
Hi-Res 5 (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3205.JPG)
Hi-Res 6 (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3206.JPG)
Hi-Res 7 (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3207.JPG)

5070

5071

5072

5073

5074

[Kal-EL]

Pad layout, looks like all the ground pads blew chunks.
5075

[thebanik]

Will have to check, but it seems even mine has similar burn marks on the chip.........

[Chuchnit]

Dude with LLC2 you were giving it 1.7x vcore at load.

[Witchdoctor]

Damm that sucks ....

Is this common to the UD7 or the chips themselves,

I have been around 1.64 to 1.665 with mine.

[Chuchnit]

The only other "incident" I know of with any board is a pre-release UD7 and ES 2600k that miahallen had.

[Kal-EL]

Chip 2 does x56 :D

Heres what chip two looks like.

llc1
1.66vCore
1.20vtt
1.72vDimm
Still on bios F2

Hi-res (http://www.overclockingalliance.com/Kal/P67_UD7_SOCKETBURN/100_3209.JPG)
5081

[DOM]

looks like rma time.... can newegg replace them ?

but ill keep the x56 lol

[Kal-EL]

Damm that sucks ....

Is this common to the UD7 or the chips themselves,

I have been around 1.64 to 1.665 with mine.
I don't think its common but peeps should peep the underside to be sure.

looks like rma time.... can newegg replace them ?

but ill keep the x56 lol
Its still a good x55 chip tho so I'll keep it for an oca duder after I test it on legitmate sub-zero cooling.

Lemme know if you guys find any sorta abnormalities on the underside of your chips. :good:

[freeagent]

Yikes, that really sucks :(

Sorry to see, but I do have to admit it looks pretty neat :)

[OC Maximus]

Well thats sucky on the first chip. Looks like the board could be the culprit. Hope your better chip doesn't get fraked up too. When I pulled my chip out a few days ago there were no signs of burns and I've wPrimed it at 1.7v a hand full of times. I use LCC LV1 only though.

[Vivi]

Sheeeet!!

My 1st cpu that i did 1.8v with :P had slight burn marks. same with my 2nd cpu, they al burn on the same spot for me (middle top right'ish) when looking at kal-el's photos.

What is causing this? high volts and too little contact between pins?

kal's burns are bad tho! 3rd degree!

[OC Maximus]

I'd try this:

Max Voltages used:
vcore 1.67
vdimm 1.72
vtt 1.22
level 2 LLC

Fully load the CPU and check voltage at vcore readout point. Should be around the same place on your board me thinks.

Then try Level 1 LLC and see what you get.

Then try a better bios :D

[Kal-EL]

Thx OCM, too late, covered in nail polish, foam and eraser atm. If it crashes and burns, I'll know I've given too much vcore. :D

[OC Maximus]

Wingin' it like a true Superhero :good:

[Chuchnit]

iirc, raja said something about intel lowering the number of pins dedicated to vtt being one possible culprit. It looks like the pads have good contact marks on the cpu. Shit just keeps getting flakier and flakier with intel.

[Neuromancer]

No burn marks on my chips, but I never got past 52X

[Witchdoctor]

Don't even want to look at mine.... pushing 1.8 for 5.6 ... total fail :ohcrap:

[Kal-EL]

Chip #2 bugs out at -105c in the os @ x56.
changing mulit-s past -65 thru -105c will not set any multi, no matter, but will run if pre-set at -50ish
Cascade is too cold basically is what I'm saying :P

[Splave]

No burns here on any chips, even latest at 1.75v. Maybe a gigabutt thing? What kinda vtt you using kal, dont tell me 1.5v

[Kal-EL]

No burns here on any chips, even latest at 1.75v. Maybe a gigabutt thing? What kinda vtt you using kal, dont tell me 1.5v

1.5v not standard for intel chips? WTF?!!?!?!?!?


1.21 and 1.22 highest.

[Splave]

Does it really need that? I get no added stability from 1.15v+

[Neuromancer]

really? I need 1.25+ to get my chip to max out :S

I have not hit 1.3 yet.. but close..


Guess you were making up for it with tranistory voltages by upping vdimm .2 out of spec :)

[Kal-EL]

Does it really need that? I get no added stability from 1.15v+

It seemed to need it when raising the bclck to 1002.9

[Witchdoctor]

I have been at 1.25 but seems to make little difference so have since backed to 1.15 with no effect on stability ....

Can't break through to 5.6 no matter what I do... :Dizzy:

tryied 1.68 core yesterday .... no help ... :shock:

[Vivi]

I have been at 1.25 but seems to make little difference so have since backed to 1.15 with no effect on stability ....

Can't break through to 5.6 no matter what I do... :Dizzy:

tryied 1.68 core yesterday .... no help ... :shock:

but have you tried 1.8 ? ;)








dont tho

[Witchdoctor]

1.8 ??? :scared:

I am using an H70 :laughing:

Have not taken it cold ............... yet

[OC Maximus]

I'm using 1.20-1.24 qpi/vtt from 5.54Ghz - 5.56Ghz.

Without it she won't boot. Maybe its a gigabyte board thing????

[Witchdoctor]

That didn't help me at all yesterday trying to boot at 5.6 ....

5.3 is all I can get out of it

[DOM]

im using 1.1v vtt x56 on ss and 1.1v for x55 on water

[GeorgeStorm]

Blimey, bit worrying :/
Atelast it's a decent chip and it's still going ;)
Even better with your 2nd chip
Will be running mine under SS hopefully, can't afford to kill it :P

[Neuromancer]

im using 1.1v vtt x56 on ss and 1.1v for x55 on water

Nice! 1.1 is stock right?
Mine BSODs until I up it to 1.2+

I can run stock up to about 4.5GHz then it starts needing more..

Oh.. I BCLK mine though... ~103-104

[DOM]

i think stock is 1.050v

i ran 106 with 1.14v

mine doesnt like to boot with more then 1.18 idk why but been like that on 3 chips

i can add more in windows though

[Splave]

Dom does your chip start with an L or a 3?

[DOM]

L

[Kal-EL]

3

[OC Maximus]

L

Edit: Oops, I'm not Dom :P

[DOM]

must be that salt water air :taunt: lol

[Witchdoctor]

Mines an L

[Chuchnit]

Mine starts with a 3 too. Plus I'm on the UD7 and I think my chip degraded ~100mhz last night. I'm still at work or I'd be testing to make sure that's the issue.

[DOM]

im at work to :(

but mine seems to be fine so far been using 1.6v-core lvl 5, 1.1vtt, mem 1.64v 1600mhz :laughing:

rest on auto = stock

but i havent checked the pads since i put it on the ss

but its used cpu and was used on a ud7 and pads have pin marks and was a lil discolored but not burned :)

2 i sold didnt have any marks or pads burns etc.

[kikicoco1334]

that sux man burned your chip :(

i am kinda glad i waited finding out everything bad about it :P

[topdog]

Looks like I had the same problem with my first 2600k, both chip and board still worked fine

http://img.photobucket.com/albums/v133/scoobytopdog/P1040234-1.jpg
http://img.photobucket.com/albums/v133/scoobytopdog/P1040240.jpg
http://img.photobucket.com/albums/v133/scoobytopdog/P1040242.jpg

[Kal-EL]

Looks like I had the same problem with my first 2600k, both chip and board still worked fine

http://img.photobucket.com/albums/v133/scoobytopdog/P1040234-1.jpg
http://img.photobucket.com/albums/v133/scoobytopdog/P1040240.jpg
http://img.photobucket.com/albums/v133/scoobytopdog/P1040242.jpg

Yah, looks the same, any ideas as to the cause?
Board/bios?

[topdog]

It's a UD7, I dont know what really caused this but it's down to one of three things in my mind

1. When I first got it, I ran it on the stock cpu cooler and when playing around with different multipliers and voltages temps got a bit high and could have scorched the cpu, the socket does look like it's melted a bit.

2. Condensation due to phase cooling

3. The same problem that caused yours to burn ;)

Voltages used
vcore 1.62
vdimm 1.76
vtt 1.14
level 2 LLC

Max clocks 5.28ghz

I have replaced the cpu with another and used it for a while on similar settings with no signs of burns on it

[Kal-EL]

Kool, thx for sharin, I'm thinking it has sumtin to do with the bios and LLC settings. Of course, I'm just guessin since I haven't hooked up any DMM's to check if theres major voltage swing offs.

You're lookin good on the submissions too bro ;)

[kikicoco1334]

when i saw the pix i though it was condensation too due to the second pix there...

[Splave]

you guys think maybe too much pressure on one side of the chip could be causing this with our non stock bolt down coolers?

no acceptable for such an expensive board imo

[Neuromancer]

If it was coolers it would be happening on other boards.

There was a great thread about hte sockets used. It gets confusing because what is written on the hold down plate is NOT who makes the socket. So one company was getting blamed when in fact they don't make the sockets causing the issue..


Should be easy to find, think it was sin's posts over at OCN about miahs post that talked about it. I looked the stuff up myself and it is true. There are different manufacturers involved. only one with problems, and only GB uses them

EDIT: LOL I dont know why I thought the valuable info was posted at OCN heh heh wasnt there.

Thing is lotes makes sockets. they also make just the hold downs. There are 3 companys that produce sockets (Lotes foxconn and cant think of the other) Gigabyte uses the LOTES hold downs like ASUS but does not use the lotes socket.

thread had the part numbers involved and everything it was very informative.


It is not a "gigabyte issue" except that it is only a gigabyte issue because they are the only one using the combination that fails.

[Vivi]

sucks tho cause a UD7 is so expensive. I would have hoped they changed sockets for the B3 rev.

I know they did this last time, quickly changed sockets on their 1156 boards when they heard the problem.

On a side note, despite all the hideous socket burns, i have never heard of someone's cpu not working afterwards. They always still work fine :O

[Chuchnit]

Rich isn't that just for s1156? Kal and Miahallen are the only two I can think of for reporting socket burn on s1155. Still sucks that it's the same vendor. I still gotta pull mine out and check.

[Chuchnit]

Pulled my cpu out of the socket yesterday and can report no signs of burn here. Must be lvl 2 llc or some condensation that did it.

[Kal-EL]

bios?

[Chuchnit]

I have used f2, f3c and f3h

[Kal-EL]

I'm 99% sure the P67AUD7B3 has a bad case of anal glaucoma. This board's bios's suck random smurf dung.

anybody got a favorite bios to try? I tried most all of them and they are all squirly like Rocky.

[Splave]

i think its the dual bios that makes things screwy, sometimes you fail to the second bios which is probably a different version, then that fails and your back to the original causing massive headaches

[OC Maximus]

I always flash the backup bios with what ever the primary bios is. Alleviates a lot of headaches.

Alt+F12 within 1 sec of white giga screen on boot up after cmos has been cleared.

[Splave]

I have a theory, ud7 sucks the gold off your chip base then kills itself so you have to rma. They then extract the gold and send you another board to repeat the process

[OC Maximus]

Its a very lucrative business :laughing:

[Chuchnit]

I always flash the backup bios with what ever the primary bios is. Alleviates a lot of headaches.

Alt+F12 within 1 sec of white giga screen on boot up after cmos has been cleared.

:thumbsup: Do the same here and have a smooth experience with the UD7. :)


Kal I haven't tried the F3's yet but I like F2c the best out of the F2 series. What are they on this week..... F4?? Been too darn busy to look.

[Vivi]

ROFL SPLAVE!

got my ud7 and,
Yea to be honest the rebooting errors piss meh off. the restart button hardly ever RESTARTS the thing. Always causes some other crap to happen.

its basically identical to UD4

[Kal-EL]

I have a theory, ud7 sucks the gold off your chip base then kills itself so you have to rma. They then extract the gold and send you another board to repeat the process

Suckleware? :blink: