[H] delided a Kaby Lake i7 7700K, the results are .. interesting

[Drak3]

11 minutes ago, Lays said:

Because 20-30c drop is worth the 1-2 year lifespan of the liquid metal that can be replaced with 2 minutes of work

Yeah, but that's like, 2 whole minutes. As in 120 seconds, or 1/30th of an hour, or 0.0000038 years.

 

/s

[Lays]

1 minute ago, Drak3 said:

Yeah, but that's like, 2 whole minutes. As in 120 seconds, or 1/30th of an hour, or 0.0000038 years.

 

/s

same

[Bleedingyamato]

8 minutes ago, Ryan_Vickers said:

I guess they'd trade lasting forever (down to a few years) for improving load temps by 20+ degrees

I'd prefer not having to perform what seems to equate to open-heart surgery on an expensive CPU any more than absolutely necessary.  If I was considering deliding which I'm not.  

 

8 minutes ago, patrickjp93 said:

There are a number of non-conducting thermal pastes which fit this image, but most people tend to choose the "max temp reduction NAO!" approach.

It hardly seems worth the extra effort having to keep replacing it over and over IMO.  

 

7 minutes ago, Lays said:

Because 20-30c drop is worth the 1-2 year lifespan of the liquid metal that can be replaced with 2 minutes of work

2 minutes and a horror film level of anxiety over "does it still work?!"

 

There's less risky ways to get lower temps I assume.  Or at least somewhat less risky than deliding every 2 years so long as you keep using the CPU in question.  Like water cooling for people braver than me who are ok with liquids in their computer.  Yes I know it's plenty safe when done properly but one leak and there goes the farm.  To iffy for my liking..  ?

[Lays]

1 minute ago, Bleedingyamato said:

I'd prefer not having to perform what seems to equate to open-heart surgery on an expensive CPU any more than absolutely necessary.  If I was considering deliding which I'm not.  

 

It hardly seems worth the extra effort having to keep replacing it over and over IMO.  

 

2 minutes and a horror film level of anxiety over "does it still work?!"

 

There's less risky ways to get lower temps I assume.  Or at least somewhat less risky than deliding every 2 years so long as you keep using the CPU in question.  Like water cooling for people braver than me who are ok with liquids in their computer.  Yes I know it's plenty safe when done properly but one leak and there goes the farm.  To iffy for my liking..  ?

 

You don't have to re-seal the CPU.  I've been using non re-sealed chips for over a year now

[MageTank]

10 minutes ago, Bleedingyamato said:

I'd prefer not having to perform what seems to equate to open-heart surgery on an expensive CPU any more than absolutely necessary.  If I was considering deliding which I'm not.  

 

It hardly seems worth the extra effort having to keep replacing it over and over IMO.  

 

2 minutes and a horror film level of anxiety over "does it still work?!"

 

There's less risky ways to get lower temps I assume.  Or at least somewhat less risky than deliding every 2 years so long as you keep using the CPU in question.  Like water cooling for people braver than me who are ok with liquids in their computer.  Yes I know it's plenty safe when done properly but one leak and there goes the farm.  To iffy for my liking..  ?

You can mitigate the risks. Using a delid tool is by far the easiest when it comes to the delidding itself. Once the processor is delidded, cover any contacts near the die with tape, or clearcoat nail polish, and you will not have to worry about getting any CLU (or your paste of choice) on the contacts. I personally do not reseal my delidded chips, because there is no point to it. Your motherboards socket retention arm will hold the IHS on to the CPU, and it's still very easy to get the CPU out of the socket, even with the IHS not being attached.

 

Dow Corning is not a bad paste though, and is chosen because of it's longevity. I have yet to find a paste that is rated for the same amount of thermal cycles, and it's very inexpensive. I believe Intel uses TC-5688 these days, but I could be wrong. Either way, you can't really go wrong as long as you clean up the glue, and ensure a nice flat even coat on the die and IHS. If you do chose to go with CLU, I recommend slightly sanding the underside of the IHS a little. CLU does not bond to nickel well at all, and it's far easier to spread on the copper underneath the nickel coating. You can spread it on the nickel though, just be prepared for hours of frustration. Oh, and don't use CLU on the top side of the IHS, it's really not worth it, and will absolutely destroy any aluminum it comes in contact with.

[Bleedingyamato]

 

21 minutes ago, MageTank said:

You can mitigate the risks. Using a delid tool is by far the easiest when it comes to the delidding itself. Once the processor is delidded, cover any contacts near the die with tape, or clearcoat nail polish, and you will not have to worry about getting any CLU (or your paste of choice) on the contacts. I personally do not reseal my delidded chips, because there is no point to it. Your motherboards socket retention arm will hold the IHS on to the CPU, and it's still very easy to get the CPU out of the socket, even with the IHS not being attached.

 

Dow Corning is not a bad paste though, and is chosen because of it's longevity. I have yet to find a paste that is rated for the same amount of thermal cycles, and it's very inexpensive. I believe Intel uses TC-5688 these days, but I could be wrong. Either way, you can't really go wrong as long as you clean up the glue, and ensure a nice flat even coat on the die and IHS. If you do chose to go with CLU, I recommend slightly sanding the underside of the IHS a little. CLU does not bond to nickel well at all, and it's far easier to spread on the copper underneath the nickel coating. You can spread it on the nickel though, just be prepared for hours of frustration. Oh, and don't use CLU on the top side of the IHS, it's really not worth it, and will absolutely destroy any aluminum it comes in contact with.

Good to know but I'm not currently considering deliding.  

 

I've got a Cryorig C7 on my 6700K (no OC) and even running RE7 at max settings it's in the mid to upper 50's C.  I have a H7 I'm waiting to install so once that's in place it'll keep my 6700K even cooler.

 

So even if I wanted to delid there's really no need to take the risk since my CPU isn't getting hot enough to really need it.  

[Lays]

17 minutes ago, Bleedingyamato said:

 

Good to know but I'm not currently considering deliding.  

 

I've got a Cryorig C7 on my 6700K (no OC) and even running RE7 at max settings it's in the mid to upper 50's C.  I have a H7 I'm waiting to install so once that's in place it'll keep my 6700K even cooler.

 

So even if I wanted to delid there's really no need to take the risk since my CPU isn't getting hot enough to really need it.  

 

RE7 = resident evil 7?

For reference, with a delid + 5.2 ghz, 1.424v load, I'm sitting in the mid 50's in every game I play at max settings

The power of delid is strong

[Bleedingyamato]

21 minutes ago, Lays said:

 

RE7 = resident evil 7?

For reference, with a delid + 5.2 ghz, 1.424v load, I'm sitting in the mid 50's in every game I play at max settings

The power of delid is strong

Yeah.  It seems to be pushing my graphics card more than most other games I've personally played.  It's got my G1 1070 at 79% load and 58C atm.  It was running the GPU fans about 40% so I locked to that with afterburner so avoid annoying sound from periodic changes to fan speed.  

 

 

It may be but the power of the anxiety attempting a delid would cause me would be equally strong.  ?

[PRTI]

8 hours ago, DarkBlade2117 said:

When they're all 75C and it is 83-85C kind of annoys me

OCD?

[zMeul]

 

[ttam]

People compare Linus... He is kinda amateur in this kind of thing.

[zMeul]

 

[MageTank]

5 minutes ago, zMeul said:

 

For the love of all that is holy, do not use CLU like this guy did. He is in for a world of hurt once that gigantic amount of CLU leaks out on his contacts. CLU bonds to gold and will short that thing out. You only need to slightly dip your bristles into the CLU to paint the die, and once more to paint the underside of the IHS. CLU will bond to itself and form a perfect seal. It only needs to be razor thin.

[Elerek]

Wouldn't the whole % reduction in temps thing make more sense if it was:

ambient temp: 21.1C

pre operation cpu temp: X

post operation cpu temp: Y

 

percent reduction = 100 - { 100 * ( Y-21.1) / (X-21.1) } %

 

?

 

basically base "0" off room temp so that your cpu temps reflect accurately how heat dissipation actually works.

[zMeul]

3 hours ago, MageTank said:

For the love of all that is holy, do not use CLU like this guy did. He is in for a world of hurt once that gigantic amount of CLU leaks out on his contacts. CLU bonds to gold and will short that thing out. You only need to slightly dip your bristles into the CLU to paint the die, and once more to paint the underside of the IHS. CLU will bond to itself and form a perfect seal. It only needs to be razor thin.

you can very well use nail polish to cover the die side exposed contacts

[MageTank]

1 minute ago, zMeul said:

you can very well use nail polish to cover the die side exposed contacts

I know, I did that myself. He did not, lol. His second attempt was much better though, but still needs to be smoother. I also highly recommend painting a thin square layer on the IHS (sand it slightly, gallium doesn't bond well to nickel, but bonds nicely to copper). Having a thin layer on the die, and a thin layer on the IHS allowed a perfect bond to form (gallium bonds perfectly to itself) and made my temps 23C cooler on my 6700k. Also saves CLU, as it's quite pricey depending on where you get it.