3770k IHS Removals - CPU temp dropped from 79C to 71C

[tscompusa]

Btw guys I did some tests with the EK Supreme-HF using direct die cooling you can see them here.

1.36v ,4.8Ghz ~60ºC runing prime.

nice results. the different mounts are just trying different pressure on the die?
i am gonna try to do direct today with mx-4 and will report back.

as long as my raystorm block allows me to get a good enough contact.

[Darknight]

Liquid Pro applied in die and IHS 4.5ghz and the temps dropped to 10 ° C compared to having mx-4.

Now I have 3770k - 4.5Ghz @ 1.16V with 58 º C full linx AVX

[tscompusa]

I just did direct die with mx-4. keep in mind this is directly after placing the mx-4 on the die and no time to cure..
when i let mx-4 cure about 24-48hrs the temps drop another 2-3c atleast from my experience.

Ambient temps are 31C

Using the Raystorm water block, its so easy to get good / safe tension on the cpu without using the IHS.

I am not sure if i can tighten it down more or not but i got a 7-8C drop from mx-4 + mx-4 vs now direct die + mx-4 (no heat spreader).

Voltage is currently set high at 1.335

Running prime95 with Large FFT

This is with 48 hrs of the mx-4 being placed (30c ambient temps)



This is with 5minutes of the mx-4 being placed (31c ambient temps)

[tscompusa]

Liquid Pro applied in die and IHS 4.5ghz and the temps dropped to 10 ° C compared to having mx-4.

Now I have 3770k - 4.5Ghz @ 1.16V with 58 º C full linx AVX

you never tried direct die yet? im curious what id get vs liquid pro with IHS vs direct die right now.. im thinking i wont see any improvement.

But if I put the liquid pro inbetween the die and waterblock i bet id go down even more.. but i dont know if i want that liquid pro stuff touching
my waterblock after what i read about it.. i dont want to have to lap my waterblock anytime soon.

btw how are you able to run so low of voltage? i tried 1.225 and got the blue screen.

I just ran 1.280 just now and seems ok though. temps went much lower also.

[andressergio]

decided to do it and man awesome almost 10C

see the after and before

almost 10C !!!





BEFORE

)



AFTER

[owikh84]

decided to do it and man awesome almost 10C

see the after and before

almost 10C !!!

well done sergio
awesome mem clock 2600MHz on P67

[andressergio]

here its what im more suprised before HS removal and change paste i launched prime at that volts and i got 94C look now



*thanks a lot owi solid board huh ?

[tscompusa]

Hey Guys just thought id update.. This should help some people out.

After trying direct die with mx4 without the IHS i decided to just now switch to Liquid Pro inbetween the die and IHS and PK1 inbetween the heatsink and IHS

And wow.. big difference! Do not do direct die on the CPU if you have the same water block I do.. it must be concave and bowed it doesnt cool for crap..

I Just dropped another 10c+ across the board.. (and paste is not cured yet for sure - been only 10min)

Id like to thank Babalouj for the recommendation with this method .. it is the best method by far!

And thanks to everyone else for finding this issue with the cooling. Now my water cooling setup is actually working the way it should be.

Ambient temps are only 2c different from before (28c) / with higher vcore also. 1.3v

[emertX]

here its what im more suprised before HS removal and change paste i launched prime at that volts and i got 94C look now



*thanks a lot owi solid board huh ?

Which paste have you used and how?
I mean with or without IHS?

[C-N]

Hi all ,

I just got round to pulling the IHS off my 3570K



I got -6oC under prime load just replacing the Intel TIM with Arctic Silver 5 between Die & IHS & IHS to Antec 620. I have a few more pastes including some Liquid Ultra to try later.

This drop is not quite so much as most however my IHS was fitted very close to the PCB & I needed to break a razor up to get thin enough blades to get in there. Even with the razor blades it was still a struggle.

[andressergio]

Which paste have you used and how?
I mean with or without IHS?

Hey !

i used artic MX2 and its with IHS on, some ppl dont put it

[andressergio]

Hi all ,

I just got round to pulling the IHS off my 3570K



I got -6oC under prime load just replacing the Intel TIM with Arctic Silver 5 between Die & IHS & IHS to Antec 620. I have a few more pastes including some Liquid Ultra to try later.

This drop is not quite so much as most however my IHS was fitted very close to the PCB & I needed to break a razor up to get thin enough blades to get in there. Even with the razor blades it was still a struggle.

C-N why u used artic 5 :/ use MX4 or MX2 will drop much more

[C-N]

C-N why u used artic 5 :/ use MX4 or MX2 will drop much more

I came home from the Pub feeling all brave & thought right lets get this done so jumped right in there only to later find out I didnt have any MX2/4.....

I'm going to try PK-1 or the Gelid stuff next then try some Liquid Ultra. When I'm happy I plan on re-fitting the IHS.

[andressergio]

I came home from the Pub feeling all brave & thought right lets get this done so jumped right in there only to later find out I didnt have any MX2/4.....

I'm going to try PK-1 or the Gelid stuff next then try some Liquid Ultra. When I'm happy I plan on re-fitting the IHS.

ahahahaha lol well next time come brave but use a good paste lol

[babalouj]

Hey Guys just thought id update.. This should help some people out.

After trying direct die with mx4 without the IHS i decided to just now switch to Liquid Pro inbetween the die and IHS and PK1 inbetween the heatsink and IHS

And wow.. big difference! Do not do direct die on the CPU if you have the same water block I do.. it must be concave and bowed it doesnt cool for crap..

I Just dropped another 10c+ across the board.. (and paste is not cured yet for sure - been only 10min)

Id like to thank Babalouj for the recommendation with this method .. it is the best method by far!

And thanks to everyone else for finding this issue with the cooling. Now my water cooling setup is actually working the way it should be.

Ambient temps are only 2c different from before (28c) / with higher vcore also. 1.3v

Nice! Now you are getting almost the same temps as me with my exact same setup (liquid pro/ultra, pk1, and raystorm). I also found that the raystorm does not like direct die contact because of the bow.

[theseeker]

I had been debating between SB & IB due to the heat issues with IB. Thanks to the OP and everyone else for sharing great information. I will get the 3770k.

[Chukundur]

Most of the pastes get dry and curing under the IHS very soon after the application and the results getting close to the original TIM.

[trance565]

Okay, I'm trying yo do this, hopefully to get atleast even temp distribution across all 4 cores. However, I can't seem to get the blade in anywhere NEAR as easy as you did in the video, am I doing something wrong here?

Edit: got it

[babalouj]

Most of the pastes get dry and curing under the IHS very soon after the application and the results getting close to the original TIM.

That is not true, most of the higher end TIMs will last at least a couple of years before "drying out". Also, Liquid pro/ultra will not dry out period.

[trance565]

wow, I used AS5 that i've had for a good 4+ years (IHS and heatsink) and got waaaay better temps, normally i'd be in the 70-76 range by now (about 10 mins into prime) and it's sitting at around 60, 66 max. still have core 0 running way cooler than the rest, but i guess he's just a cool guy

update: 30 minutes in, and 70* is the max, compared to a usual 78-81

intel should fix this

[WhiteFireDragon]

In the video, I'm watching that blade and PcCI2iminal's fingers as he's slicing through the IC, thinking: "oh gawd I hope that blade just doesn't slip or snap". That could go right into your hands/fingers.

[Chukundur]

That is not true, most of the higher end TIMs will last at least a couple of years before "drying out".

Yes, but not when pastes are between the die and the IHS with a poor contact between them. I already tried МХ-3, NТ-H1 and TR CFIII. Beginnings results were very good, but within 1-2 weeks temperauturite dramatically increase, especially the last core. I never tried Liquid pro/ultra. Let other users who used silver paste to share how long the results are preserved in this case.

[st0ned]

I've been using a LM paste for over a week. Temperature is what they used to be.

[Allubz]

Yes, but not when pastes are between the die and the IHS with a poor contact between them. I already tried МХ-3, NТ-H1 and TR CFIII.

Why would this make a difference?

[babalouj]

Yes, but not when pastes are between the die and the IHS with a poor contact between them. I already tried МХ-3, NТ-H1 and TR CFIII. Beginnings results were very good, but within 1-2 weeks temperauturite dramatically increase, especially the last core. I never tried Liquid pro/ultra. Let other users who used silver paste to share how long the results are preserved in this case.

It sounds like you have poor contact between the IHS and the core because you did not scrape the black silicone/sealant off the perimeter of the chip and IHS. My temps have not changed since I first applied my liquid ultra.