3770K owners tempted to delid

[Fudd Rucker]

Sorry if this OT , I remember a friend back in the day playing with his Barton core 2500+ mobile chip and he accidently set the voltage to almost 6v on an msi k7n-delta2 lsr. What a show that was lol.

[Phill]

So today arrived my Coolaboratory Liquid Ultra from Ireland. I will be using it between my IHS and the cpu die. I will post any improvements in temp.

[GT7R]

Originally Posted by Phill

So today arrived my Coolaboratory Liquid Ultra from Ireland. I will be using it between my IHS and the cpu die. I will post any improvements in temp.

That is exactly the one I've used for my 3570k. It's under a fanless NH-D14 and maxes 78 in prime @4.6 - 14 degrees better than with the stock TIM.

[Phill]

Today I decided to replace the MX-2 between the IHS and CPU die. In addition I decided to glue the IHS back to the CPU. Here we go!

First some load temps taken after an 80min of stress testing (pay no attention to the elapsed time showing 16 seconds).


Next I applied the CLU AND RTV sealant.


Here the CPU socket is firmly holding the IHS to the cpu


After some time the H110 was seated to the cpu using MX-2 between the H110 block and the IHS


CPU now stressed again for 80 min (ignore elapsed time)



Results

The AVG temp on all cores dropped 7.775C
The average MAX temp on all cores dropped 6.75C

Thoughts
Success I would say. I am pleased with the results. I am a little concerned that I poorly applied the MX-2 after the relidding was complete. Also seeing as this was my first time applying Coolaboratory Liquid Ultra I hope I did not apply too much of it on the cpu die.

Time will tell if the RTV sealant holds but I am quietly confident it will.

To recap, delidding, cleaning Intels crap TIM and reapplying MX-2 TIM achieved a temp drop of 15C.
Replacing the MX-2 between the IHS and die dropped avg load temps another ~8C.
In total that is an improvement of 23C.

[KaptCrunch]

Originally Posted by Phill

Today I decided to replace the MX-2 between the IHS and CPU die. In addition I decided to glue the IHS back to the CPU. Here we go!



Next I applied the CLU AND RTV sealant.




After some time the H110 was seated to the cpu using MX-2 between the H110 block and the IHS


CPU now stressed again for 80 min (ignore elapsed time)



Results

The AVG temp on all cores dropped 7.775C
The average MAX temp on all cores dropped 6.75C

Thoughts
Success I would say. I am pleased with the results. I am a little concerned that I poorly applied the MX-2 after the relidding was complete. Also seeing as this was my first time applying Coolaboratory Liquid Ultra I hope I did not apply too much of it on the cpu die.

Time will tell if the RTV sealant holds but I am quietly confident it will.

To recap, delidding, cleaning Intels crap TIM and reapplying MX-2 TIM achieved a temp drop of 15C.
Replacing the MX-2 between the IHS and die dropped avg load temps another ~8C.
In total that is an improvement of 23C.

good job bring down the temps, the RTV shouldn't be 100% circle needs a gap so the lid can breath

through hole and not balloon IHS

[nasrott]

so it sounds like it is worth it to gain the extra degrees from them using the TIM, so besides the paste what is this other stuff ya need? RTV at an automotive store, is it just gasket maker?

[Phill]

Yes AFAIK it is a gasket maker. I have had previous experience with my particular RTV and thought that it would be perfect for relidding. I also used a syringe to apply the RTV. I could not figure out how else to apply it and end up with a half decent job.

I am now considering delidding, CLU'ing and relidding my other 3770k and 3570k...

[nasrott]

so no one has come up with a retail copper shim to use direct die for all this? Still think that would give the best temps no?

[Evil Genius Jr.]

[TPU] EK Water Blocks Announces De-Lidded Ivy Bridge CPU Cooling Solutions

[nasrott]

so they stop on the mounts instead of pressuring the pcb it seems