GIGABYTE's mobo X79-UD3 happens self-ignition and burning!

[sin0822]

UD7 and the UD3, UD5 and Assassin 2 share the same exact PWM, and use the same MOSFETs. THey use them in different quantity, the UD7 has 16 phases, thus the load balance is spread over more MOSFETs.

We should hear a statement later today from GB about this, at least for the uSA.

[Warwian]

UD7 and the UD3, UD5 and Assassin 2 share the same exact PWM, and use the same MOSFETs. THey use them in different quantity, the UD7 has 16 phases, thus the load balance is spread over more MOSFETs.

We should hear a statement later today from GB about this, at least for the uSA.

Right, just like I thought than!

I'm eagerly waiting to hear what they have to say, as this probably isn't going to do wonders for sales of their new "3D power"-motherboards. Gigabyte X79 Ultra Burnable Motherboards? :p On a more serious note however, I like Gigabyte and I trust that they will handle this situation in the best possible manner, I doubt many other companies would go as far as recalling the products for example.

[dctokyo]

I only read BIOS recall, am i alone or what? I don't see board recall.

The bios update was released but it essentially cripples your board so you can't overclock or overvolt at all. Only other thing is replacement

Gigabyte offered two remedies. First, it issued a BIOS update for existing owners, which prevents this from happening. It does so, however, by thottling the CPU when it's under extreme stress, making these boards practically useless to overclockers, and is only recommended for those who run their setups with mild overclocking that isn't voltage-assisted. The second remedy, of course, is an unconditional return for replacement. Overclockers can send their burnt boards, as well as healthy ones that are prone to this issue, for a free replacement, when corrected boards are available.

[zsamz_]

The bios update was released but it essentially cripples your board so you can't overclock or overvolt at all. Only other thing is replacement

than what is it good for

[Hondacity]

hahahahah

[sin0822]

I am pretty sure that 99% of what wanted to be said was lost in translation, so those sites that just translated and made statements like board's recalled are a bit overreaching. Cripple BIOS, lol that is a good one since it really doesn't do anything but stop users from setting switching frequency and vcore response high enough to kill their boards. It doesn't change OCP and OTP limits, and your OC wont throttle any more than it did before.

[zsamz_]

hahahahah

is it my birthday?

[TurboDiv]

, when corrected boards are available.

So they will be replacing these garbage PWMs ? Crippled BIOS is such a bad way to remedy things...

[zsamz_]

So they will be replacing these garbage PWMs ? Crippled BIOS is such a bad way to remedy things...

hope people not gonna wait 3 months like i did when the p67 b2 boards were recalled

only thing they can do is cripple the bios for now

[dctokyo]

than what is it good for

GIGABYTE is going to host the first Under-Clocking Championship's with it it's the new trend

[Falkentyne]

I think some people are owed a BIG Apology for being flamed and accused of being incompetent overclockers. Seriously, blaming the USER for a hardware flaw or substandard parts? No wonder Lee's sig talks about overclockers doing into mode here...

And the recall is official. This is not going to sit pretty for Gigabyte.
Their P67 boards were very nice and decent quality, only problem were the BIOS bugs and boot loop issues (Hello? where's the onboard 50 cent Chinese made BIOS switch, huh?), but the hardware is fine. They still need to *FIX* the erratic LLC vcore swings with that multi step LLC crap though...F3d/f3g/f3h on the UD4/5/7 had VERY stable and predictable LLC...all they had to do was lessen the vrise under LLC2 at 1.4+v, and it would have been absolutely *perfect.* Instead they up everything with the multi step LLC, which must have only been tested at low vcore settings (since multi step LLC actually works **very** nice at close to stock voltages).

[ice_chill]

I think some people are owed a BIG Apology for being flamed and accused of being incompetent overclockers. Seriously, blaming the USER for a hardware flaw or substandard parts? No wonder Lee's sig talks about overclockers doing into mode here...

And the recall is official. This is not going to sit pretty for Gigabyte.
Their P67 boards were very nice and decent quality, only problem were the BIOS bugs and boot loop issues (Hello? where's the onboard 50 cent Chinese made BIOS switch, huh?), but the hardware is fine. They still need to *FIX* the erratic LLC vcore swings with that multi step LLC crap though...F3d/f3g/f3h on the UD4/5/7 had VERY stable and predictable LLC...all they had to do was lessen the vrise under LLC2 at 1.4+v, and it would have been absolutely *perfect.* Instead they up everything with the multi step LLC, which must have only been tested at low vcore settings (since multi step LLC actually works **very** nice at close to stock voltages).

Give me any socket 2011 board with 6core/12thread highly overclocked watercooled CPU, running IntelBurnTest with BIOS setting that put extreme stress on the VRMs and no air flow to cool the components, and I will kill any board....honest

Ofcourse it was a user error.

[Hondacity]

active cooling on r4e...

[Falkentyne]

Give me any socket 2011 board with 6core/12thread highly overclocked watercooled CPU, running IntelBurnTest with BIOS setting that put extreme stress on the VRMs and no air flow to cool the components, and I will kill any board....honest

Ofcourse it was a user error.

Too bad I don't have a "Not sure if Serious" Joker pic to upload...
Are you trolling?

The OP wasn't putting his CPU under those conditions...
I love how you completely twisted everything I said.
Nowhere did I say extreme overclocking and nowhere did I mention Linpack, now did I?

Or maybe Gigabyte's recall is all fake?

Someone give me a tinfoil hat, please.
I actually think you're trolling.

[Vicey]

Not on a Mobo box, but I remember I had ASUS P5NT Deluxe socket 775, it came with an additional clip on fan for the VRM heatsink, it had a warning that the fan needs to be installed if water cooling is used.

But another point is Mobo makers don't need to say that, it says in the stress testing tools that excess stress will be applied and components can be damaged, a warning like that should make you research what it's about and how to stress test safely.

I had the same thing with my P6T6 X58 board. It had a fan which clipped on to the PWM Heatsink and the slip inside the box said it should be used if water cooling or overclocking for maximum stability.

Personally I think the problem presented in the video is 50/50 - Gigabyte should have shut the system down earlier so that the board didn't catch fire and this guy should have fitted a fan to the PWM heatsink. I'm not pro overclocker but even I know leaving a board on a test bench without airflow over the PWM heatsinks is a bad idea.

[ice_chill]

Too bad I don't have a "Not sure if Serious" Joker pic to upload...
Are you trolling?

The OP wasn't putting his CPU under those conditions...
I love how you completely twisted everything I said.
Nowhere did I say extreme overclocking and nowhere did I mention Linpack, now did I?

Or maybe Gigabyte's recall is all fake?

Someone give me a tinfoil hat, please.
I actually think you're trolling.

You got my point wrong, from the video you see a watercooled system, then you see him loading prime95, which is used for stability testing, so water cooling + stability tests makes me assume the system is overclocked otherwise there would be no point to everything he did.
Also using water cooling instead of air cooling makes me assume the system was overclocked quite well.

I replied to you because you said it is Gigabytes fault, I disagree because the act of overclocking alone voids the warranty.

[Falkentyne]

If it were his fault, then how come there is a major recall? I mean this is pretty BIG. And how come Gigabyte replaced his motherboard and gave him a new extreme edition CPU?
Believe me, all this stuff with the new Bios deliberately limiting the boards from overclocking, and this recall directly prove that there's a VRM problem with these boards. I'm waiting for you to explain this recall.

Plus as I also mentioned--which you also failed to address, there have been users who have had the VRM's fry RIGHT after turning on the boards.
AND
furtermore, gigabyte, and others, said that it was a defect on the board that caused this failure. Yet you're accusing the user. You really need to do your research and read more, before jumping to conclusions.

[ice_chill]

I did my research and went for ASUS

[BeepBeep2]

Too bad I don't have a "Not sure if Serious" Joker pic to upload...
Are you trolling?

The OP wasn't putting his CPU under those conditions...
I love how you completely twisted everything I said.
Nowhere did I say extreme overclocking and nowhere did I mention Linpack, now did I?

Or maybe Gigabyte's recall is all fake?

Someone give me a tinfoil hat, please.
I actually think you're trolling.

[STEvil]

You got my point wrong, from the video you see a watercooled system, then you see him loading prime95, which is used for stability testing, so water cooling + stability tests makes me assume the system is overclocked otherwise there would be no point to everything he did.
Also using water cooling instead of air cooling makes me assume the system was overclocked quite well.

I replied to you because you said it is Gigabytes fault, I disagree because the act of overclocking alone voids the warranty.

With how quickly the FET burned up a fan obviously would not have stopped what happened.

[Dainas]

Gigabyte has gone massively down hill the in the last few years, where as Asus seems to have cleaned up their act a touch. The whole 1156 socket debacle where Gigabyte tried to hide foxconn sockets under lotes branded brackets that was it for me. I'm still not a fan of Asus though, knowing them they are just savvier with their cost cutting.

[STEvil]

They havent cleaned up the Z7S issue..

[sin0822]

If it were his fault, then how come there is a major recall? I mean this is pretty BIG. And how come Gigabyte replaced his motherboard and gave him a new extreme edition CPU?
Believe me, all this stuff with the new Bios deliberately limiting the boards from overclocking, and this recall directly prove that there's a VRM problem with these boards. I'm waiting for you to explain this recall.

Plus as I also mentioned--which you also failed to address, there have been users who have had the VRM's fry RIGHT after turning on the boards.
AND
furtermore, gigabyte, and others, said that it was a defect on the board that caused this failure. Yet you're accusing the user. You really need to do your research and read more, before jumping to conclusions.

There is no major recall, the translation from Chinese to English used the word recall,and some sites went to town on it, and added in their own commentary. Here is how it stands in the USA:
http://www.gigabyte.us/press-center/....aspx?nid=-869

Might shed some more light on things.

[cx-ray]

That's how it stands in Europe as well. They state it's due to a BIOS error that doesn't enable OVP when overclocking. Additionally, the affected boards are backed with a 5 year warranty.

[JohnZS]

This is crazy, is it me or has this been the year of PWM's?

First we had rumours of GTX 570 cards having weak PWM's then followed the 590 cards and their epic PWM failings and now this?!?
I suppose one could also add that this is going to put a dent into Gigabyte and the X79's reputation when it is not really entirely Gigabyte's fault that they were shipped poor PWMs, however they should have tested this as it seems rather crazy that this is happening.

I have also heard that some Bulldozer boards are burning out too

John