Does IHS delete help temps?

[lexidobe]

My 4790K runs hot enough to almost throttle at 5ghz and 1.18 volts. Even at stock speed undervolted it ran at 91 degrees even cooled by a custom water cooling loop. I did a hammer and vice delid, which did not make any difference. I lapped my CPU and waterblock which helped a little. My question is if I completely remove the IHS, will it run cooler without having the thermal resistance of a heat spreader and another layer of thermal compound. My thinking is the IHS has the same amount of surface area to touch the IHS anyway, and that the waterblock will cool it better if it is directly touching the die than if it touches something that touches the die. Any thoughts?

[0ld_Chicken]

My CPU runs hot enough to almost throttle at 5ghz and 1.18 volts. Even at stock speed undervolted it ran at 91 degrees even cooled by a custom water cooling loop. I did a hammer and vice delid, which did not make any difference. I lapped my CPU and waterblock which helped a little. My question is if I completely remove the IHS, will it run cooler without having the thermal resistance of a heat spreader and another layer of thermal compound. My thinking is the IHS has the same amount of surface area to touch the IHS anyway, and that the waterblock will cool it better if it is directly touching the die than if it touches something that touches the die. Any thoughts?

I htink most people get a couple °c off their load temps when running bare die compared to a good delid with liquid metal TIM.  What TIM did you use when you delidded? and what processor are you running 5ghz at 1.18v?

[NinjaJc01]

My 4790K runs hot enough to almost throttle at 5ghz and 1.18 volts. Even at stock speed undervolted it ran at 91 degrees even cooled by a custom water cooling loop. I did a hammer and vice delid, which did not make any difference. I lapped my CPU and waterblock which helped a little. My question is if I completely remove the IHS, will it run cooler without having the thermal resistance of a heat spreader and another layer of thermal compound. My thinking is the IHS has the same amount of surface area to touch the IHS anyway, and that the waterblock will cool it better if it is directly touching the die than if it touches something that touches the die. Any thoughts?

5GHz is overkill.

[NinjaJc01]

I htink most people get a couple °c off their load temps when running bare die compared to a good delid with liquid metal TIM.  What TIM did you use when you delidded? and what processor are you running 5ghz at 1.18v?

i7-4790k

[0ld_Chicken]

i7-4790k

stock speed, undervolted, it ran 91°c on a custom loop?  thats not right at all and you've got some serious cooling issues.  I hit 80c while running 1.365

edit: and you must be the only person on earth able to run 5ghz at 1.18v so congrats there

edit: my bad didn't realize you werent OP

[lexidobe]

I used Gelid GC extreme thermal compound. If the IHS is not needed, why does it have one?

[TheRandomness]

I used Gelid GC extreme thermal compound. If the IHS is not needed, why does it have one?

Because it's harder to get good contact between the CPU and the cooler without it. Just saying, 90 with a custom loop!? Is your pump even working?

[lexidobe]

Because it's harder to get good contact between the CPU and the cooler without it. Just saying, 90 with a custom loop!? Is your pump even working?

Pump works. I guess I just have a hot running CPU.

[0ld_Chicken]

I used Gelid GC extreme thermal compound. If the IHS is not needed, why does it have one?

its a ton easier to mount with a IHS and it helps spread the heat, hence the name, heat spreader.  Its best to use a liquid metal TIM between the die and IHS, other TIM dont really make a difference.  Get some cool labs liquid ultra/pro. 

and you must be the only person on earth able to run 5ghz at 1.18v on 4790k so congrats there

[TheRandomness]

Pump works. I guess I just have a hot running CPU.

No, it should top out at 80. 90 and you probably have a problem somewhere in your loop.

[0ld_Chicken]

Pump works. I guess I just have a hot running CPU.

or mount problems, or something else, anything other than just a hot cpu.  NO 4790k will run at 90c at stock speed, undervolted, on a well functioning custom loop

[Guest]

5GHz is overkill.

*looks away*

[NinjaJc01]

*looks away*

How?

[Guest]

How?

*looks away*

[lexidobe]

Any ideas?

[0ld_Chicken]

5.0 is life

ok, 4988MHz.... whatever

[0ld_Chicken]

Any ideas?

removing the IHS will do you no good at all if you get results like you were saying before. you need to figure out where your problem lies first.  I do know that most TIM (like gellid) "pumps out" from between the die and IHS so maybe that is what you are experiencing.  I'd try a liquid TIM between the die and IHS like I said before.  they are the only ones (besides maybe kryonaut) that do not pump out and lose effectiveness 

[Lays]

5 ghz 1.18v lol 

 

If you're hitting 90c+ at 1.18v, you're doing something wrong, or your voltage isn't 1.18v.  I only hit 60c with a custom loop under load at 1.33v 5 ghz.