Does anyone have problems with Asus P67?"

[MaxAwesome]

DISCLAIMER:

READ CAREFULLY IF YOU USE AN ASUS P67 Motherboard.

First things first: This thread is NOT meant to scare any ASUS P67 users, but rather to inform them of what happened to me TWICE in 10 months, and let them know that we might do best to exercise caution when using these motherboards, WHETHER we overclock or not.


Now, to the "story", I will try to present in a timeline sort of way, so it's better to understand, and not just a wall of text.



MY HARDWARE AT THE TIME OF FIRST DEATH:

CPU: i7 2600k @ 4.2 | 1.224v LOAD.

MOBO: ASUS P8P67 Evo (non-B3)

PSU: NOX Urano 700w

RAM: 4GB |2x2gb| Kingston HyperX 1600mhz @ 1.65v

GPU: Sapphire HD 6870 1GB DDR5.




TIMELINE OF EVENTS - 1st DEATH (P8P67 Evo + 2600K) Thread on overclock.net about it.

• Late January 2011: I upgrade to Sandy Bridge. I buy a P67 Evo, a 2600K and 4GB of Kingston HyperX 1600 1.65 (more on this later);
  
  
• From January until February 11th, all works fine and dandy. Sig rig sits at a comfortable and perfectly safe OC @ 4.2ghz with 1.224v under MAX LOAD. All other voltages on AUTO, and RAM was running @ 1.55v 1600mhz (they were 1.65v, but stable at 1.55v);
  
  
• February 11th: My computer just shuts down while playing CS 1.6. Out of the blue. Never to turn on again. Symptoms: PC won't turn on at all, fans spin for a fraction of a second when power button is pressed. Nothing else happens.

• All parts are taken to the store, where the motherboard is deemed dead.
  
  
• I still kept my first 2600K, because no one, not even me, suspected of anything else other than a faulty mobo AFTER THE PSU was tested and deemed OK. BTW, that PSU still works on my secondary PC since february. 100%.

• February-March: During this time, I was without a system until late March, when the B3 Revision mobos started rolling out. I still had my netbook though, but it was a pain to use as a main for 2 months.
  
  
• Fast forward to late March, (30th, I believe), my new B3 EVO arrives...
  
  
• ... Only to discover that the previous P67 EVO had killed the CPU as well. I couldn't find a board to test the CPU on during the recall. Bear that in mind. So I only found out my 2600k was dead when my new mobo came.
  [*]
  
• Late march: my 2600K goes through RMA, I got a new one in about a week. All is good again.
  
  
• Early April: I buy a BRAND NEW Corsair TX650, just to be extra cautios. Might as well invest on a decent PSU once and for all, right?

IMPORTANT REMINDER: This 1st death occured during the painstakingly long Intel mobo chipset recall. I was without a new motherboard for almost 2 months.



Now, we move on to the second death....



MY HARDWARE AT THE TIME OF SECOND DEATH:

CPU: i7 2600k @ 4.2ghz | 1.248v LOAD

MOBO: ASUS P8P67 Evo B3

PSU: Corsair TX650w

RAM: 8GB |2x2gb| +|2x2gb| Kingston HyperX 1600mhz @ 1.65v

GPU: Sapphire HD 6870 1GB DDR5.


TIMELINE OF EVENTS - 2nd DEATH (P8P67 Evo B3 + 2600K) My thread here on Techpowerup about it. And on Xtremesystems too.

• From Late March / Early April, everything is absolutely fine and dandy. Sig Rig working wonderfully without a single hiccup.
  
  
• Fast forward to late October: BAM! Dies again. This time, not during CS 1.6, but during a 3Dmark Vantage run. Pc shuts down in the same manner as the first time, out of nowhere. Fans do the typical split second spin, and then nothing. Only death.
  
  
• Take PC to my store again: ANOTHER DEAD MOBO and CPU.
  
  
• In the meantime, I swap my RAM to 1.5v DIMMs and with XMP profiles. My previous HyperX ones didn't.

• Right Now (November 17th): Sig Rig has been working fine, for now. ALL is at stock now.
  
  
• I won't touch anything else on this motherboard.

MY CURRENT HARDWARE:

CPU: i7 2600k @ STOCK.

MOBO: ASUS P8P67 Evo B3, EVERYTHING @ stock.

PSU: Corsair TX650w

RAM: 8GB |4x2gb| G.Skill RipJawsX 1600mhz @ 1.5v XMP

GPU: Sapphire HD 6870 1GB DDR5.


BOTTOM LINE:

1st DEATH: In less than a month, my first P67 Evo dies and takes the CPU with it.

2nd DEATH: This time, it took 6 months to die... but it still died.

Will it happen a third time? I don't know, but I'm afraid, to be completely honest.


One last thing, and this is, IMO, the most important thing:


I just couldn't let this go. I KNEW I wasn't doing anything wrong and yet, my hardware had died on me TWICE, under the same mysterious circumstances.

I consider myself an experienced Overclocker, and while at first I though "Oh crap, what am I doing wrong... I hardly believe it's me afterall. It must be something in the hardware.


So I decided to digg in as deep as I could... and while I'm NOT listing EVERY case I found, I'm posting the most relevant ones.

And...

...It's not happening to just me. In case you were wondering if I'm must be in fact doing something wrong, well, there's been a few "deaths" scattered around the web.

The problem is exactly the same: PC shuts down, won't turn on again. in the end, DEAD MOBO and CPU.



Compiled list of ASUS P67 deaths, that also killed the CPU: (Some weren't even overclocking, others were gaming, others were idle).


http://vip.asus.com/forum/view.aspx...7&id=20110426021716988&page=1&SLanguage=en-us

http://hardforum.com/showthread.php?t=1593874

http://www.benchtec.co.uk/forums/threads/8115-P8P67-Rev3-Woes.......

http://adrenaline.uol.com.br/forum/.../369578-asus-p8p67-placa-mae-bomba-fujam.html

http://www.overclock.net/t/1014337/dead-p8p67-pro-18-days-old

http://www.overclock.net/t/1042932/p8p67-dead-question-about-upgrade

http://communities.intel.com/thread/20868


PART II


http://www.xtremesystems.org/forums...ssion-thread&p=4816984&viewfull=1#post4816984

http://www.xtremesystems.org/forums...ssion-thread&p=4817920&viewfull=1#post4817920

http://forums.overclockers.co.uk/showthread.php?p=18903994

http://www.phforums.co.za/f14/i72600k-asus-p8p67-dying-63481/

https://vip.asus.com/forum/view.asp...board_id=1&model=P8P67&page=1&SLanguage=en-us

http://www.xtremesystems.org/forums...ssion-thread&p=4970457&viewfull=1#post4970457


And... You know what the Icing on the Cake is?

Just Google : "p8p67 dead", if you have google suggestion ON, you will see google automatically filling in your search.

See what it does? There must be a WHOLE lot of more cases, for Google to have it on auto-suggest already.


CONCLUSION

In the end guys, this is NOT a thread against ASUS. To be honest, they are by far, my only choice for motherboards and I love ASUS stuff. I have 2 notebooks and netbooks from them only., but after seeing my stuff die for no apparent reason, I'm scared.

 

[NdMk2o1o]

Sucks to have had 2 boards and CPU's die though there is still nothing that lays the blame squarely on the boards and especially as you are not just singling out anyone board but Asus P67 boards in general.

I feel your pain, but I am sure if one looked hard enough I could also find the same with other brands of boards and perhaps multiple instances, it doesn't mean that wholse series of boards is no good, it means sometimes shit happens and it also happens to other people just to show us that we really aren't as unlucky or songled out as we feel sometimes...

Just my 2cents though.

 

[bogmali]

Good info. I ran my P8P67 EVO for 3 months folding 24/7 with a 2600K and never had anything die on me. My OC was also @ 4.2 but occasionally it will error out on the folding clients but this is mainly memory related. Yeah that blows that you had to go through 2 cpu's and mobo's.

 

[mjkmike]

I would realy like to see the back of those dead chips. Sounds to me that the memory controller that is built into sandy bridge got zapped. TPU's own cadaveca has said allot about running memory on sandy above 1.5 and XMP profiles. Take some time and read a few revews of his. P.S my P8P67 has been crunching hard at 100% at 4.5 for months and the P8p67 pro crunched for over 6 months at 4.8 untill I gave that rig to a friend.

 

[cadaveca]

A few factors that could have contributed to the death:


"PLL OVERVOLTAGE" set to enabled.

"IMC Voltage(VCCIO)" @ 1.2v or higher

Ram voltage over 1.5v on a kit not designed for SandyBridge(every ram kit is made for specific platforms. Yes, they might work in another platform, but at the same time, if it wasn't on the board's QVL, then you are SOL".




That said, I do have a P8P67 pro, and an extra 2600K. Both are sitting in thier boxes.

 

[de.das.dude]

i got an rma'd board with me now and my VRM has is on the verge of death again. asus service here is really pathetic. im going gigabyte or msi after this.

 

[ChristTheGreat]

this must be bad luck, or bad batch of motherboard... I run mine for almost one month, 4,4ghz 1.256v at 100% CPU all day on BOINC.

This can happen, and it's sad . If there were too much dead mobo, I think Asus would recall the board.

Like mjkmike said, I also saw about ram, that'S why I bought on newegg Kingston HyperX genesis Sandy bridge edition. I think also this could be a memory controller problem with 1.65v. Some people said they had 1.5v, but like I said, no body is protected from hardware failure. But I wish mine won't die haha.

Thanks for your feedback MaxAwesome

A few factors that could have contributed to the death:


"PLL OVERVOLTAGE" set to enabled.

I wasn't able to get over 45x, since new bios, enable this function, and I can get higher. But this can let you overvolt the PLL, but I don't think this could cause killing the motherboard.. that would be nice: THis function will kill your motherboard, please keep on disable.. that doesn't make sens I think..

 

[cadaveca]

I wasn't able to get over 45x, since new bios, enable this function, and I can get higher. But this can let you overvolt the PLL, but I don't think this could cause killing the motherboard.. that would be nice: THis function will kill your motherboard, please keep on disable.. that doesn't make sens I think..

I've had three CPUs degrade in a matter of minutes...with PLL overvoltage enabled, and 1.38v. Two different boards.


I always base my comments by my own personal experience. It's very odd that both CPU and board die, but it's also pretty common for Sandybridge for both to go at the same time, so while it may seem a bit off that there is no warning for this BIOS option, I do not feel it is needed for there to be a warning, as overclocking, period, can affect stability and lifespan of components.

CPUs are of varying quality. Some will handle certain votlages fine, while others will not. You always take a chance when overclocking, and everyone already knows this, so I don't see why it wouldn't make sense. To me, it's no different than X58 and running 1.65v ram, while needing to keep VTT within 0.5v of vDIMM. Not many boards tell you about this, but at this point, it's common knowledge.

 

[Wrigleyvillain]

He also said all other voltages left "on AUTO" which is actually NOT safe when overclocking cause you don't know what it's doing with them. And the monitoring software often not very accurate even if you try to pay attention.

 

[_JP_]

I want to know more of this, because I was thinking of buying a P67 board.

i got an rma'd board with me now and my VRM has is on the verge of death again. asus service here is really pathetic. im going gigabyte or msi after this.

Did you happen to notice that your board has a 4 phase VRM for the CPU with no heatsinks at all...I wouldn't say ASUS is bad...you're just going beyond what it can do and therefore the board is failing on you.

 

[ChristTheGreat]

I've had three CPUs degrade in a matter of minutes...with PLL overvoltage enabled, and 1.38v. Two different boards.

Please let me know how you know your CPU degraded.

Anyway I saw alot of people, when they reach max multiplier, beeing able to get more with Pll overvoltage enable..

I was able to boot at 46x but with 2001 bios, I need it enable..

 

[cadaveca]

Please let me know how you know your CPU degraded.

Anyway I saw alot of people, when they reach max multiplier, beeing able to get more with Pll overvoltage enable..

I was able to boot at 46x but with 2001 bios, I need it enable..

The CPUs all needed more voltage for the same clocks after 10 passes of LinPack "maximum memory" load with PLL Overvolt enabled.

And yes, PLL overvolt will enable extra multipliers, but really this is intended for those using extreme cooling methods, not for daily usage, in my opinion.

This option is not available on all boards, either.

My review hardware gets pounded, on purpose, as I do try to find the failing point so that I can report it to you guys, as well as the OEMs who products I review.

 

[John Doe]

The first one went away because it was "Non-B3". On the second one, you probably had bad luck.

You really can Google any mobo with "dead" added next to it, and find applicable results to your situation. It has no backing at all. Also, that Nox PSU could make a potential failure point. I read up on those before (I think they were rebadged SuperFlower's {old}) but can't remember right now. If it's a crap unit, it might just shot the mobo down from bad VSB regulation.

Further, the EVO is based on the same PCB as the P67 Deluxe. So it's a solid board like the rest of Asus` Sandy offerings. It's always nice to have opinionated paranoia in a forum... in one way or another.

 

[cadaveca]

It's always nice to have opinionated paranoia in a forum... in one way or another.

Unless you work for one of these companies that got his dead hardware, and you've tested to confirm how the failure occurred, there's no way ANYONE can say what really happened here. Myself included.

What I can say, however, is that deaths of 1155 hardware in the same sort of manner is not something ASUS-specific. Gigabyte, MSI and others all have some reports of deaths like this.

 

[Millennium]

I am curious if anyone else gives credence to running 1.65v memory causing this? Because I have been running it on my rig, on and off, due to problems with my lower voltage memory and sheer lazyness. Should I get that stuff out of my system pronto?

By the way cadaveca I have been running (as far as I know) PLL overvoltage and 1.4v ish @ 4.5 cause I have a crap CPU and I haven't noticed any degredation, but your experience worries me. I'll check my settings when I get home.

 

[John Doe]

Unless you work for one of these companies that got his dead hardware, and you've tested to confirm how the failure occurred, there's no way ANYONE can say what really happened here. Myself included.

Re-read my post, he doesn't know why they exactly failed either.

Asus users don't have a "ticking bomb" in their hand and that's a fact. All this thread does is to panic people for no reason.

 

[cadaveca]

By the way cadaveca I have been running (as far as I know) PLL overvoltage and 1.4v ish @ 4.5 cause I have a crap CPU and I haven't noticed any degredation, but your experience worries me. I'll check my settings when I get home.

I am also running my current personal CPU with PLL overvoltage enabled as well. The CPU has degraded a bit, but then stopped. I really think just some CPUs can handle it, and some cannot, and when the CPU goes, it blows out to ground, and this kills the board at the same time.


It's also worth mentioning that improper cooler installation can cause death as well, if it happens to tweak the socket pins the wrong way. there's no way for any of us to know what happened in this particular situation, but based on my experience and reports on the internet, I feel confident that this problem is not an ASUS-only issue.

Of course, not many people realize that the PLL chip is also chip on the board itself, either. On the board in question, there is THIS chip, which supplies clocks to PCIe.:

The option in BIOS refers to an INTERNAL PLL on the CPU. Not many people know exactly what enavbling the option does, as the actaul PLL votlage does not change when this optoin is enabled. If it does adjust the PLL for PCIe, it might make sense that this was causing the issues.

 

[de.das.dude]

I want to know more of this, because I was thinking of buying a P67 board.

Did you happen to notice that your board has a 4 phase VRM for the CPU with no heatsinks at all...I wouldn't say ASUS is bad...you're just going beyond what it can do and therefore the board is failing on you.

dude. i used to do 3600 24*7 on my orignal, with this crap now i cant even OC 200mhz :/

 

[cadaveca]

dude. i used to do 3600 24*7 on my orignal, with this crap now i cant even OC 200mhz :/

When you don't have an 1155 rig, you comments do not directly relate to anything but your own rig. You run AMD, not Intel.

 

[jsfitz54]

@ cadaveca: If you run stock only, will there be the kind of failures listed here?

I am waiting for a 2600k and asus p8p67 deluxe to be delivered from another TPU member. I purchased 2x4gig 2133mhz gskill to match with system and it's 1.65v stock...is this a poor/bad match and should I stick with 1.5v memory exclusively?

 

[MaxAwesome]

Guys, first of all thanks for ALL the input.

But, I'd like you to know, that while I know I made the title of the thread a bit sensationalist, my goal is NOT to diss ASUS (as I have said twice in my OP, and I've been using and trusting ASUS for a long time now).

Besides, this is not only about the motherboards dying, but also the fact that they take the CPU with them.

A dying mobo is nothing to write home about. But when it starts taking CPUs with it, with a fairly significant number of cases, then I think it's worth at least documenting.

And after long searches, this issue seems to happen "only" or mostly to ASUS P67 users.

I want to try to help others who have gone through the same thing, because even today, I don't have a clue what happened.

My research only lead me to the only obvious common factor in always failures like mine: The ASUS mobo.

That's all guys. I don't want panic, I don't want mistrust on ASUS.

I've changed the title of the thread to something less alarming.

 

[cadaveca]

@ cadaveca: If you run stock only, will there be the kind of failures listed here?

At stock, no, you should not have any issues at all, really. Of course, as always, I suggest that nothing is left to auto, whether overclocked or not.


Also, anything over 1333 MHz memory and 1.5v is overclocking. Sandybridge supports 1333 MHz max.

http://ark.intel.com/products/52214/Intel-Core-i7-2600K-Processor-(8M-Cache-3_40-GHz)

http://ark.intel.com/products/52210

I make the same recommendations to anyone...run stock, no PLL overvolt, and no more than 1333 MHz 1.5v ram. Anything more than that is taking chances. Not always a big chance, but as always, overclocking does have it's risks, and as silicon processes get smaller and smaller, they get more and more fragile. If you overclock, and something dies, I do not feel that you should RMA anything, too. Abusing the RMA system is not something I'll ever recommend, as it does make parts more expensive than they need to be.

 

[MaxAwesome]

Cadaveca, my previous Kingston 4GB Kits, were 1.65v kits, and while they were not declared "SB approved on the box", I did find them on the Kingston SB compatibility list, but not on ASUS..

That's when I switched to G.Skill RJX 1.5v, with XMP.

is XMP a good thing or a bad thing? I have the impression that Intel recommends it? Not sure.

Because I really DO hope that what caused the death of my HW was my previous RAM. Even though I ran it at only 1.55v. But still, you never know.

At least this Ripjaws kit I have now is SB-approved, and made especially for it.

 

[jsfitz54]

I make the same recommendations to anyone...run stock, no PLL overvolt, and no more than 1333 MHz 1.5v ram. Anything more than that is taking chances. Not always a big chance, but as always, overclocking does have it's risks, and as silicon processes get smaller and smaller, they get more and more fragile. If you overclock, and something dies, I do not feel that you should RMA anything, too. Abusing the RMA system is not something I'll ever recommend, as it does make parts more expensive than they need to be.

I respect your input but by the same token the Manufaturers sell, "enthusiast = made for overclocking" motherboards.

These types of "enthusiast" mobo's should fair better as they are promoted to used better base materials in the build, example: ASUS Sabertooth Tough Military Grade.

Too the same end, Intel and AMD pitch the same assertions.

Getting what you pay for is not unfair.

is XMP a good thing or a bad thing

XMP was invented by Intel.

 

[MaxAwesome]

XMP was invented by Intel.

Ok, thanks