i5 Overclocking causes pin burnouts

[bobsmith1492]

I was just diving into overclocking my new i5 setup; I Primed it overnight at 150Bclk and was going to up it again today when I saw the new article on Anandtech:

1156 Sockets Burning Out at High OCd Loads

I have a Gigabyte GA-P55-UD4P so I'm sure my board is affected.

Now I wonder; how far should I really push it? This sucks... I was looking forward to some serious tweaking on this thing.

Edit:

Cliffs -
On Intel 1156 sockets, overclocking beyond about 4.2GHz has been causing the contact points between the CPU and socket to burn up on motherboards made with Foxconn sockets, apparently because of insufficient contact force or misalignment between the pins and the CPU pads.

 

[Zensal]

It says it only affects serious overclocks. Like the suicide run kind. It says they have been running an i5 @ 4.3GHz with no problems.

 

[Bill Brasky]

I'm a little concerned this might affect long term stability. Some think this problem will eventually arise even at stock and that overvolting just exposes the flaw earlier.

 

[Jumpem]

Read this thread at hardforums: http://www.hardforum.com/showthread.php?t=1460140, particularly posts by lathode.

The pin prick indentations are there from the factory. They are there on brand new, never inserted CPUs. This has been verified via photos in the linked thread.

It appears that the motherboard socket leaves longer dash looking indentations, which are evident on every pad in the AT photos.

 

[imported_Lothar]

Originally posted by: Jumpem
Read this thread at hardforums: http://www.hardforum.com/showthread.php?t=1460140, particularly posts by lathode.

The pin prick indentations are there from the factory. They are there on brand new, never inserted CPUs. This has been verified via photos in the linked thread.

It appears that the motherboard socket leaves longer dash looking indentations, which are evident on every pad in the AT photos.

I read that entire thread.
Lathode may be on to something but he didn't answer this question...
How come there are NO burn outs on a LOTES socket so far? Pure luck? Skillz? Did "Merlin the Wizard" put on his robe and wizard hat?

When I see people with the LOTES socket experiencing the same problem, I will stop blamming Foxconn.
Until then, I won't.

For someone like you who only plans running on stock, it's a non-issue.
For someone like me who wants to do 4.00GHz flat, it's an issue.

This issue I just read about the past 2 days is making me considering Socket 1366 again, but I still have about a month to decide.
If I go the Socket 1156 route, I will be picking up an eVGA board at Micro center.
I will keep returning by claiming "Product is faulty, it doesn't recognize RAM" until I get one with a LOTES socket.

If I go the Socket 1366 route, I will be picking up an ASRock X58 Extreme.

 

[AznAnarchy99]

fuck really... I just got my 1156 and I looked at the board and its a foxconn. It is wayyy to much work to remove everything, mobo and all, and the megahelem. Im kinda regretting not spending the extra $150 for an 1366...

Should I be worried going for 4ghz?

 

[RussianSensation]

Originally posted by: AznAnarchy99
fuck really... I just got my 1156 and I looked at the board and its a foxconn. It is wayyy to much work to remove everything, mobo and all, and the megahelem. Im kinda regretting not spending the extra $150 for an 1366...

Should I be worried going for 4ghz?

Don't worry about it Azn. The problem likely wont' occur until way into 4.2-4.3ghz zone. Also don't forget the issue arises once you start pumping massive voltages (1.5-1.65V) as was highlighted in Xstremesystems forums. You won't be running 1.5V on air cooling 24/7. Right now my rig is 100% stable at 3.9ghz at only 1.300V (bios) / 1.312V in CPU-Z. You simply cannot compare the power consumption of a Core i7 system at these voltage levels vs. 1.55V @ 4.3ghz. So for non-extreme overclockers, these issues are likely irrelevant.

 

[jaggerwild]

I been over 4.6 EVGA 758 A1 i noticed the Foxconn logo on the socket when I got it new. Gonna redo the thermal compound take a peek in it......... I have noticed a few things that are not normal, but with frequent bios updates thought it may be attributed to that.

 

[yh125d]

I'd go for whatever clock you're looking for. Make sure its stable and whatnot, but after a couple weeks check the socket. If it looks any different from normal at all, back off a tad

 

[imported_Lothar]

Originally posted by: RussianSensation

Originally posted by: AznAnarchy99
fuck really... I just got my 1156 and I looked at the board and its a foxconn. It is wayyy to much work to remove everything, mobo and all, and the megahelem. Im kinda regretting not spending the extra $150 for an 1366...

Should I be worried going for 4ghz?

Don't worry about it Azn. The problem likely wont' occur until way into 4.2-4.3ghz zone. Also don't forget the issue arises once you start pumping massive voltages (1.5-1.65V) as was highlighted in Xstremesystems forums. You won't be running 1.5V on air cooling 24/7. Right now my rig is 100% stable at 3.9ghz at only 1.300V (bios) / 1.312V in CPU-Z. You simply cannot compare the power consumption of a Core i7 system at these voltage levels vs. 1.55V @ 4.3ghz. So for non-extreme overclockers, these issues are likely irrelevant.

Some experienced(not Newegg reviewers, but HardOCP and other tech forums) people have been having problems with stability at stock settings, or even booting.
Couldn't this be attributed to the problem?

Also, could this issue have any relation to the 3 motherboards blowing up in Tom's review, or was that something completely different?

I really hope this whole thing turns out to be a facade and is not something one should be concerned about with a 4GHz constant overclock.
I really don't feel like paying an arm and a leg for what I think is an overpriced eVGA board.

 

[Borealis7]

While i do agree this is a serious matter and should be addressed by the M/B manufacturers, i still think that if you overclock you cant hold anyone but yourself responsible for damages.

 

[krose]

My i7 860 on a Ggiabyte P55-UD3R would not pass stability tests at stock settings. I tested everything and could not find a cause. The odd thing is that it was fine in single channel but dual channel would fail. Both sticks of RAM were fine in single channel. After I read the article about the Foxconn sockets I removed the CPU and found that a significant number of pins were not making contact. So IMO it can affect stock setups.

 

[Borealis7]

that's ok, but if it were a problem occurring on overclocked settings, then it is not an issue the M/B makers should address. in fact, they can capitalize on the issue by making "special overclock" boards that fix this.