LM isn't just Gallium, there are other metals in there to make it a more suitable TIM. I can't imagine anyone has ever tested for this so it's hard to say what the long term effects might be.
Personally I'd just wipe it off then clean with alcohol and start over to be safe but that's just me. I'm sure it's fine though.
You can tell if you use too much by holding it vertically, the excess will roll down. You'd know if you used too little because there would be exposed die, when I get done with mine it looks like square metal puddle.


The water (if any is still there) would evaporate and escape (assuming you didn't seal around the entire IHS).

 

I read to use just enough that you can't see the die and not so much that there's any pools that have a ripple effect when going over them. The die of the 6700k is so shiny and is virtually the same color as the liquid metal, but everything is running fine. I did use black RTV silicone around the entire IHS though... :worriedsm

 

Not certain about the moisture. If you left a breather hole when you applied the RTV, any moisture should evaporate rather quickly.

Did your power requirements drop, or just temps?

 

Yeeeaaah, I didn't... :worriedsm

Just temperatures. Temperature wasn't a bottleneck for my overclock... I just didn't want to go over 1.35V (setting in the BIOS, not actual) and I couldn't achieve 4.5GHz at 1.35V or less (with XMP enabled that is).

 

What I've learned about liquid metals:

Conductonaut is like Coolaboratory Liquid Pro, it's more of a pure Gallinstan alloy, with a few extra metals mixed in to drop it's melt point. Liquid Ultra adds in some stuff to slow the oxidization down, and to provide a more stable bond. So the trade offs are Conductonaut and Liquid Pro have a higher max thermal transfer, but they are both more aggressive so they'll 'set', which is why you'll see a few people think either have dried out. The other downside to them when used between an IHS and cooler, is that they'll attack/alloy to unplated copper much quicker. The result being a black layer that forms, and a performance drop after 3 or so months. If you have this happen, you clean off the liquid metal, and reinstall, and it should be good for quite a while.

Gallinstan oxidizes basically instantly, that glassy shine it gets is the surface oxidizing. As long as you got the water off and didn't let it mix into the LM, it's fine.

As for bond lines, using 'Krazy Glue'(beware some of them do attack the substrate) in the corners, is a much better option, as it'll allow for a much closer mating of the core to the IHS. Apply light pressure to the IHS while either silicone or krazy glue sets to get the best 'mount'. You can also go dry mount and not worry about gluing it back on.

A note: I'm not a fan of ThermalGrizzly, because I've not had any luck getting an MSDS out of them for anything. I've read that other people have asked and TG refuses or ignores the requests.

 

You sealed it up 100%? That's weird, you a lot of the temp drop from removing the gap that it created in the first place. Removing the Intel thermal insulation helps too.


If you must seal one up RTV is fine as long as you apply it thin (only on the corners) and use the socket retainer thing to squish it before it hardens. Not sure what the super glue thing is about TBH, I'd be worried about delidding it again and having pieces of the PCB stuck to the IHS.

 

The stock IHS installations I've seen leave an area unglued. I assume it's to accommodate gas expansion/contraction due to CPU temps. Or perhaps it's merely a cost-saving measure?

Weird, all the ones I've seen were sealed all the way around from the factory including the OP's and my own.

That is odd.
Last 2 were i7-6700k and i7-7700k. Both had a 1/8" gap on one side. Didn't note factories of origin.

Haven't de-lidded the 8086. I would if I knew it would drop voltage reqs @ 5.3GHz.

Kinda kicking myself for not leaving a gap now... never thought about leaving a hole to let any moisture build-up evaporate. My 6700K DID have a gap (as you can see in the attached picture), and I've seen it in some delidding videos as well.

As far as the sealing and it's affect on temperatures, I used a VERY thin amount of black RTV silicone and spread it with my fingers, definitely less than what comes from Intel... because like you mentioned, I read that the gap this causes between the PCB and the CPU die increases heat. It's so thin in areas that maybe, just maybe, there's still a slight hole where moisture can evaporate out of. I used the Rockit 88 relidding attachment though so there was definitely pressure from the IHS onto the CPU die to make sure it sealed with minimal gap.

 

The stock IHS installations I've seen leave an area unglued. I assume it's to accommodate gas expansion/contraction due to CPU temps. Or perhaps it's merely a cost-saving measure?

 

Weird, all the ones I've seen were sealed all the way around from the factory including the OP's and my own.

 

That is odd.
Last 2 were i7-6700k and i7-7700k. Both had a 1/8" gap on one side. Didn't note factories of origin.

Haven't de-lidded the 8086. I would if I knew it would drop voltage reqs @ 5.3GHz.

 

It will, I'm not guaranteeing 5.3 but more heat means more resistance, if you've ever heard of "leakage" that's what's being referred to. If you knock 20c off the core temps you'll lower the voltage required for stability in most cases. I wasn't able to get past 4.5Ghz with reasonable voltage on my 4690k until I delidded it.

 

I'll be damned, I couldn't see it in the original pics. Maybe mine did too?
Just drill an air hole.