[SOLVED] cpu has strange white spots on bottom

[This username is unavailable11]

I was upgrading my pc and when I removed my cpu (i5 6600k) it had these strange spots on the bottom and I am not sure what caused them. The only thing that I could think of is that I had too many volts put into it when I was overclocking it to 4.4Ghz. I don't remember what voltage it was at sadly.
CPU:

https://imgur.com/SBtK3Z0

Motherboard:

https://imgur.com/OZGJV7P

 

[AllanGH]

Welcome to the world of contact fretting--the reason why sometimes just removing and replacing DIMMs clears up apparent memory problems (among other things).

Another reference.

 

[This username is unavailable11]

Welcome to the world of contact fretting--the reason why sometimes just removing and replacing DIMMs clears up apparent memory problems (among other things).

So is this a big problem to performance? Because before i upgraded it was still running fine. Also, what happened for this to occur?

 

[AllanGH]

So is this a big problem to performance? Because before i upgraded it was still running fine.

It can be.

Also, what happened for this to occur?

Did you read the references I linked to, above?

 

[This username is unavailable11]

It can be.

Did you read the references I linked to, above?

Yes I read parts of it and what I gathered is that it is caused by small movements between the connectors. What I was looking to clarify is what this unstable connection would be caused by in a PC. Would it just be a bad connection to the motherboard clamp or is it a user error.

 

[AllanGH]

With the current design of Intel processors, as well as computer systems in general, it is not possible to eliminate the incidence of contact fretting in the area you have pictured.

You CAN slow the speed of fretting by increasing the clamping pressure of the processor to the processor socket, reducing the temperature of the contact surface to 60°C, reducing the humidity of the environment, reducing the vibration within the computer case (implying a move to SSDs, eliminating fans, and moving pumps and radiators outside the case), and running the processor (and memory) at lower voltages.

An argument can be made for the application of non-conductive silicone grease to the processor socket, to lubricate the contact surfaces, but the question remains as to whether the socket materials are compatible with silicone grease. Doing so may just make the situation worse, so I wouldn't recommend it for any system that you are not willing to sacrifice to an experiment.

Processors having pins that mate into a ZIF socket are less prone to the failure mode, but are not immune to it; so, moving to a processor that uses pins will reduce the problem somewhat.

All-in-all, being aware of contact fretting, and structuring your maintenance cycle in a way that accounts for it, is probably your best move; and, it looks like you were on your way to observing some weird issues with the system, had you not decided to do some upgrade work now.

 

[AllanGH]

Let me add this:

Anytime you run into issues of crashes, memory or hard drive errors, or video 'weirdness', out of the blue, your very first action should be to open up the system and disconnect EVERY connection point and un-make and re-make the connections several times, then seat all connectors fully.

If the issue is contact fretting, just doing this will clear up the problems--for about as long an interval as before the problem first cropped-up. Better that you do this for yourself, than paying somebody else to do it for you.

 

[This username is unavailable11]

With the current design of Intel processors, as well as computer systems in general, it is not possible to eliminate the incidence of contact fretting in the area you have pictured.

You CAN slow the speed of fretting by increasing the clamping pressure of the processor to the processor socket, reducing the temperature of the contact surface to 60°C, reducing the humidity of the environment, reducing the vibration within the computer case (implying a move to SSDs, eliminating fans, and moving pumps and radiators outside the case), and running the processor (and memory) at lower voltages.

An argument can be made for the application of non-conductive silicone grease to the processor socket, to lubricate the contact surfaces, but the question remains as to whether the socket materials are compatible with silicone grease. Doing so may just make the situation worse, so I wouldn't recommend it for any system that you are not willing to sacrifice to an experiment.

Processors having pins that mate into a ZIF socket are less prone to the failure mode, but are not immune to it; so, moving to a processor that uses pins will reduce the problem somewhat.

All-in-all, being aware of contact fretting, and structuring your maintenance cycle in a way that accounts for it, is probably your best move; and, it looks like you were on your way to observing some weird issues with the system, had you not decided to do some upgrade work now.

Thank you, so basically it’s wear and tear of the system. Anything I could do to improve the fretting on the CPU or no? I was planning on selling it but not sure if that’s going to happen now.

 

[AllanGH]

::: sigh :::

Well, if it was MY system, I'd be using pure silicon grease on the contacts (and a VERY thin film of that at most) and cranking the processor engagement down as far as I could, without damaging things.

At best, it would be one of those grand experiments that I'm prone to do when I don't care if I tank things or not, and count it as a win if the system survives. Unfortunately, I have several shelves in the garage dedicated to the storage of the carcasses of systems that didn't quite make it.

I would highly recommend that, if you care about having the system operational for any reason, don't follow my example.

You could try polishing the socket contacts with a pencil eraser, and blowing out the eraser bits with canned air; but I am circumspect about trying that on a processor, because of probable static damage.

For the processor, the most I would probably try is to lay a new dollar bill flat on a hard table surface, and set the processor down on it and press it down with medium force, and push it from one end of the bill to the other--only once--then try it in the system. Dollar bill notes are actually abrasive enough to do a good job at removing the cruft from the processor.

Now that I've said what I might do.....I'm telling you now:

DON'T DO IT. You might be disappointed in the results.

Then again, if you do it and it works, you can sell the system with a free conscience.