Question about delidding

[JoshGreen89]

Hello all, I am curious about relidding with some of that black silicone gasket stuff? Should it be used, yes or no? I've been getting both answers everywhere I've been looking...
Can anyone give me an answer and their reason why or why not? Please?
Thank you very much in advance... 
Hope everyone's day is going well so far.

[Tabs]

It depends on the performance you're looking to get from the delid. The main benefit of delidding is reducing the distance between the IHS and the die (and replacing the thermal paste with something better like liquid metal).

 

Regluing the IHS to the top of the substrate using sealant will invariably add a little more z-height to the package. This will reduce the performance of the delid.

 

You'll still get a better thermal conductivity compared to stock, but it won't be as good as without resealing.

 

So it's really up to you - not re-sealing is a bit more of a hassle to work with, it's a bit more dangerous, but the potential performance gains are higher. Resealing basically gives you a better thermally conducting version of a stock cpu, and so is useful if you want better than stock performance whilst also wanting the least amount of post-delid hassle.

 

Gamers Nexus has gone into this topic in quite a lot of detail.

[airdeano]

pretty much the issue can be traced back to the IHS glue being the culprit more so than the TIM. the thickness of that glue was enough to allow a greater distance between the die and IHS. if you are keeping the CPU, no glue (just be careful if removing from mobo). if RMA or selling, then reglue.

[Neftex]

i was thinking of putting glue in corners just so it doesnt move, somewhat in between seal and no seal thermals?

[wildgg]

1 hour ago, JoshGreen89 said:

Hello all, I am curious about relidding with some of that black silicone gasket stuff? Should it be used, yes or no? I've been getting both answers everywhere I've been looking...
Can anyone give me an answer and their reason why or why not? Please?
Thank you very much in advance... 
Hope everyone's day is going well so far.

when using silicone you can create a gap between the CPU DIE and IHS and that can result in higher temps.

 

personally i move my hardware around allot (taking coolers of and cpu's out of the mobo) so i like to put small dots of silicone on the corners and let that dry for 1 hour and you're golden.

[O9B0666]

just use the cpu sockets clamp to hold it together, dont use silicone thats why you delid in the first place.

[JoshGreen89]

5 minutes ago, O9B0666 said:

just use the cpu sockets clamp to hold it together, dont use silicone thats why you delid in the first place.

Isn't the main reason to delid to change Intel's super craptastic TIM for something better, fresher even??? I may be wrong here tho...

[O9B0666]

Um yeah and I'm not saying otherwise 

[DarkSmith2]

1 hour ago, JoshGreen89 said:

Isn't the main reason to delid to change Intel's super craptastic TIM for something better, fresher even??? I may be wrong here tho...

no, to reduce the gap between IHS and DIE.

[wildgg]

i would say the main reason is 100% going from shity paste to liquid metal and not the gap..

yes closing the gap helps for sure but changing the TIM is the big factor

 

i always used 4 little dots of silicone on the corners and didn't see a temp difference between that and not using any silicone.

[JoostinOnline]

5 minutes ago, wildthing said:

i would say the main reason is 100% going from shity paste to liquid metal and not the gap..

yes closing the gap helps for sure but changing the TIM is the big factor

 

i always used 4 little dots of silicone on the corners and didn't see a temp difference between that and not using any silicone.

I 100% disagree with you there.  I did a whole topic on it explaining it (check my sig).  The included TIM isn't that bad.

[wildgg]

8 minutes ago, JoostinOnline said:

I 100% disagree with you there.  I did a whole topic on it explaining it (check my sig).  The included TIM isn't that bad.

in your topic there is no comparison what so ever or tests between mx-4 with silicon or without silicon and same goes for LM.

 

there is no way that the GAP is the big improvement here, like i said yes smaller gap helps but you realise that: mx-4=8.5 W/(mK) and Thermal Grizzly Conductonaut = 73 W/mk

 

quote from your topic:''While it definitely doesn't perform as well as higher end thermal pastes or liquid metal (just replacing the pre-applied TIM with MX-4 on a 4790K got me a 7C drop)''

so you're saying you got a 7C drop in total with the silicone removed?

[JoshGreen89]

41 minutes ago, O9B0666 said:

Um yeah and I'm not saying otherwise 

you literally said "don't use silicone that's why you delid in the first place"
sounds like otherwise to me...

 

[JoostinOnline]

3 minutes ago, wildthing said:

so you're saying you got a 7C drop in total with the silicone removed?

No, I got a 7C drop with the silicon still there.  I got around 16C (give or take a degree, it's been awhile) with it removed.  That's on an old 4790k too.

[wildgg]

2 minutes ago, JoostinOnline said:

No, I got a 7C drop with the silicon still there.  I got around 16C (give or take a degree, it's been awhile) with it removed.  That's on an old 4790k too.

gotcha, so basically your point is you don't need LM as long as you get the silicon removed and use a decent paste, but LM will give you the optimal results?

5 minutes ago, JoshGreen89 said:

you literally said "don't use silicone that's why you delid in the first place"
sounds like otherwise to me...

 

listen, this topic might become a big argument and that's fine with me as long we keep it nice

but in the end of the day if you delid the cpu adding a little bit of silicone to hold the IHS in place is not going to be the end of the world..

 

if you add a little bit of silicone and you clamp the IHS nicely back on the cpu it won't be an issue

[nick name]

The silicon is absolutely a factor in de-lid performance. If you want to re-secure you want to make certain you use the thinnest amount possible. That includes just doing the corners. If you use too much on the corners it will still seat the IHS further away from the CPU die. If it's a system that won't be moved or shifted a lot then you may not need the IHS to be glued back on. 

[JoostinOnline]

3 minutes ago, wildthing said:

gotcha, so basically your point is you don't need LM as long as you get the silicon removed and use a decent paste, but LM will give you the optimal results?

Yes.  If you're delidding, you might as well spend the extra $5 on top tier stuff.  LM doesn't degrade like some TIM does either.

[JoshGreen89]

11 minutes ago, wildthing said:

gotcha, so basically your point is you don't need LM as long as you get the silicon removed and use a decent paste, but LM will give you the optimal results?

 

11 minutes ago, wildthing said:

listen, this topic might become a big argument and that's fine with me as long we keep it nice

but in the end of the day if you delid the cpu adding a little bit of silicone to hold the IHS in place is not going to be the end of the world..

 

if you add a little bit of silicone and you clamp the IHS nicely back on the cpu it won't be an issue

 

OHHHH LOOOL I was already well aware to the infighting it would bring. I just wanted some opinions on both sides before I made my decision.
Was I not being nice? I thought so, if a bit sarcastic. But hey, it's the almighty PCMR, if not for sarcastic what would be? lol