EggXpert

I found an article over at Anand where they wrote about issues with extreme overclocking on their p55 motherboards. They found that consistently, processors did not have sufficient pin to pad contact, and as a result were damaging both the CPU and motherboards they were used in under extreme OC conditions. They mentioned that it only seems to occur in Foxconn manufactured boards and with extreme OC conditions, so I have a few questions on the topic

1. Anand say that we should be fine with mild oc // stock, but would it be possible to experience the same type of damage simply from pro-longed day to day use (such as F@H for example)

 2. The issue of pin-to-pad contact. From what I understand, insufficient contact affects the power draw, are there any other consequences arising from this? How would you typically test for this anyway? Would you just clamp down the cpu and let it sit for a while then revisit and see if there are proper markings?

3.  Has anyone experienced similar issues with their LGA-1156 setups? OC or stock?

Source of article - http://anandtech.com/mb/showdoc.aspx?i=3661

First very cool article.  I would say this is some fault with the CPU retention mechanism.  I would think it would be beyond Foxconn (I am guessing the motherboard manufacturers get them from common suppliers. Though Anand indicates Foxconn makes their own--quite possible.  And quite possible they also sell to other motherboard manufacturers!). 

It must happen from lack of proper contact between the CPU and the socket forcing a gap that electrons must jump across.   It would work just like a cold short (a break in a electrical path). There is resistance and electrical demand and a load builds up and finally jumps the gap resulting in both an arc (the carbon deposit on the pin(s) and a huge electrical surge that would then course through those delicate traces.  Probably melting them. 

Would I expect non-overclocked CPUs to behave this way?  Absolutely.  Improper clamping (unless it has to do with a third party heat-sink - Anand didn't list their I don't think) has nothing to do with overclocking.  I suspect the extreme overclocking pointed out the problem with chip retention.  I suspect over time others who are not overclocking will report similar difficulties.  Interesting.

The other thing I'm quite curious about is how the warranties would work out. If / when something like this happens, do manufacturers/retailers actually accept RMAs? Technically overclocking is done completely at your own risk so all the responsibility is on the end user. Reading this shattered any thoughts of trying to push my 6750 past 3.2. I've read tons of people going 3.6 "easily on stock voltages with g0 stepping" (which I have) but I can't get anywhere near 3.4 unless i hit the 1.425 mark, not worth it!

If it's any help, Foxconn used to make EVGA boards, but not no more, they have been making their own since 1366.

XFX, I know, why would one by one of those, are still amde by Foxconn and Jetway though, I believe, could be off on that one.

Penewah!