Any chance of fixing this delidded CPU? (i7-6700)

[sam_86314]

I ordered some computer parts on eBay that were marked as "for parts or not working". When they arrived, all of the components worked except the CPU, a delidded i7-6700. It'll boot into the BIOS, but booting into Windows causes one of two BSODs ("WHEA UNCORRECTABLE ERROR" and "SYSTEM THREAD EXCEPTION NOT HANDLED"). I'm planning on delidding it again to see if it can be repaired. I plan to just put some of my Cryorig CP7 paste on it if I can't salvage what's already on it.

The seller stated that they replaced the TIM under the IHS with Conductonaut. It looks like they used superglue to reattach the IHS.

Here are some high-res photos of the chip. Anything I should look out for when delidding it? Does anything seem off about the chip?

Spoiler: CPU Pictures

 

[Jetster]

Delid it again and try to boot direct to die. No IHS

 

[R-T-B]

Delidding a non-K cpu strikes me as something a real rookie would do...

So look for real rookie mistakes. Remember conductonaut is metal liquid and HIGHLY CONDUCTIVE. If I had to guess, that may be related.

 

[JackCarver]

When it boots into BIOS I don't think it's a conductonaut problem. So it's working for BIOS but not for windows?? Could also be a harddisk problem or RAM probem. I would first try another system, a live system from USB or DVD/CD if you get also errors there.

 

[R-T-B]

When it boots into BIOS I don't think it's a conductonaut problem. So it's working for BIOS but not for windows?? Could also be a harddisk problem or RAM probem. I would first try another system, a live system from USB or DVD/CD if you get also errors there.

I've seen oddities with conductonaut similar to this. But usually only when they scratch something else and use a metric ton of the stuff, flooding whatever surface scratch they made in the delid.

Admitedly, I have only seen this situation once.

 

[sam_86314]

When it boots into BIOS I don't think it's a conductonaut problem. So it's working for BIOS but not for windows?? Could also be a harddisk problem or RAM probem. I would first try another system, a live system from USB or DVD/CD if you get also errors there.

I ruled out drive and memory. After testing the 6700, I stuck my old 6500T back in the system, and it booted into Windows no problem. I had tried several different memory configs while testing the 6700 to no avail.

The seller of the hardware noted that he had the same problem on a different motherboard.

 

[silkstone]

Drop it in acetone for a couple of hours to losen the glue. Delid, clean it up and take a look at the die. It might be chipped for something.

 

[sam_86314]

Drop it in acetone for a couple of hours to losen the glue. Delid, clean it up and take a look at the die. It might be chipped for something.

I've been looking into this. I'd like some other opinions about submerging a CPU in acetone.

 

[biffzinker]

The super glue is going to be a pain to deal with cleaning off.

@silkstone bought a second hand 8700K that someone de-lidded, and used superglue. Might be able to offer some advice.

Edit: Nvm he posted.

 

[R-T-B]

Hmmm...

Just a thought, but around Skylake era, Intel had a unusually high number of cpus that could no reach stock clocks. Most ended up warrantied though, and the number was still low.

Maybe, as silly as it seems, just bumping voltage or lowering the multi would work, if it really is one of these "turd" chips?

 

[Flaky]

Are all the capacitors under the cpu intact?

I heard at least one success story where damaged cpu which allowed entering the bios worked much better after disabling C-states and EIST. Try that.

 

[JackCarver]

Ask the seller if it worked after delidding. If so nothing went wrong here and you can leave heatspreader at place.

And like R-T-B said try higher VCore and/or lower multi.

 

[Bones]

I've been looking into this. I'd like some other opinions about submerging a CPU in acetone.

There is nothing about the CPU acetone would do harm to, in fact may even clean the pin pads a bit and that's a good thing.
All you'd need to worry about is to be sure it's completely dry before setting it into the socket.

 

[newtekie1]

Are you booting using the integrated graphics? it could be possible that whatever was damaged, damaged the iGPU, and it works fine at 2D clocks in the BIOS, but not when the driver kicks in when Windows loads.

Judging by the scratches at the corner of the IHS, I'm guessing whoever delidded it the first time wasn't exactly gentle...

Delidding a non-K cpu strikes me as something a real rookie would do...

Looking at the job they did, I'm guessing rookie too. This might have been the "I'm going to try it on this throw away CPU before I try it on my expensive CPU".

 

[Bill_Bright]

I'm planning on delidding it again to see if it can be repaired.

Repaired? Not sure what sort of damage you believe you would be able to "see" once delidded again, let alone what could be done to "repair" said damage. I highly doubt anyone could identify the specific TIM (thermal interface material) used just by looking at it. There are 100s out there, from top shelf, sophisticated formulas from major brands, all the way down to no-name, generic, pure silicone grease. And almost all are various shades of white to dark gray. So short of "seeing" no paste at all, or some obvious foreign object stuck in there, I don't see what you expect to see.

Delidding a non-K cpu strikes me as something a real rookie would do...

So look for real rookie mistakes.

I agree with the rookie comments. But again, not sure what anyone, including a very experienced pro, might look for. No TIM? Improperly mounted lid? Foreign object? Beyond that, what?

Assuming a proper TIM was used, no foreign objects got in there, and it was all put back together and remounted properly, my guess would be ESD damage - and that cannot be seen without a very powerful (and expensive!) microscope in the hands of someone who knew what to look for.

 

[R-T-B]

I highly doubt anyone could identify the specific TIM (thermal interface material) used just by looking at it.

It's conductonaut, a liquid metal alloy. He knows this from the seller.

 

[Bill_Bright]

Oops. Sorry. I saw that, then promptly forgot it!

 

[jaggerwild]

Try booting with only a couple cores, may have killed a few off.

 

[Millennium]

Good luck. Nothing to be lost by taking the IHS off and taking a look. If there are no obvious caps missing or scratches I would leave the IHS off and try it with just thermal paste. See if you can improve the boot situation somehow (though as others are suggesting you'll be very lucky if you can get it to boot reliably now).

 

[thesmokingman]

Geeze, the guy superglued the damn IHS, so lazy!

And 2d/3rd on the use of acetone. It softens cyanoacrylate aka crazy glue.

 

[sam_86314]

While testing, I got another BSOD (MACHINE CHECK EXCEPTION). Everything points to this meaning hardware damage.

 

[thesmokingman]

Did you remove the IHS and see if it was a bad TIM job?

 

[sam_86314]

Did you remove the IHS and see if it was a bad TIM job?

Not yet. That's next on my list.

 

[Millennium]

I once knocked a pin off an Athlon 64 X2 and it wouldnt boot. I went off to buy a new CPU and motherboard, failed to do that, came back to the system later that day. I pulled everything apart, just put it back together and it worked - and continued to work for many months.

CPU problems can be hard to pin down - so my advice is don't give up unless you are certain the CPU is toast.

 

[TheMadDutchDude]

Superglue? Wow... you might have to resort to a good ol' fashioned razor blade.