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

[PcCI2iminal]

PcCI2iminal, did you re attach the IHS with glue or something or is it just loose and held together by the cpu socket clips on the mb?

nothing
just held together with the retention bracket

[Eldonko]

Cool thanks. You realize a lot of people are going to kill their CPUs trying this now right? lol

[BMWMASTER81]

The Heratkiller has no stop screws, so it was no problem for me to mount Waterblock without IHS. Only had to remove the retention bracket.

[PcCI2iminal]

Cool thanks. You realize a lot of people are going to kill their CPUs trying this now right? lol

just send the bill to Intel Corp!!
j/k

KKKKK

The Heratkiller has no stop screws, so it was no problem for me to mount Waterblock without IHS. Only had to remove the retention bracket.

i think CPU370 has no stop screws also,i will check it and i will run my naked 3770k on water,i believe its possible to prime it at 5GHz or more

here is what i got on water with same but unmoded 3770K

[Boulard83]

this is 10x harder than just removing the IHS altogether and more risky and time consuming. do not do it this way.

one slip up of your hand and you'll accidently touch the chip somewhere with the dremel, which will most probably kill the chip right away

I mean .... Remove the IHS the good way, THEN cut the IHS to leave the 4 side, put it back on so the socket hold down is still OK.

Now you can do EZ direct die cooling.

[G.Foyle]

Am I right under the assumption that the removal of the IHS and retention bracket only works with waterblocks and not air cooler as they would bend the socket pins due to their weight ?

No, it works fine with air coolers, too. Socket has plastic ridges between pins that the CPU substrate rests on and pins only exert their own pressure against lands. It's the same with IHS, only without IHS you have to distribute the pressure evenly across the whole substrate. If you only pushed against the die, the substrate might bend and not make contact with all pins, or even the die might crack.

[Sebastione]

So is it a problem at all if you take care like in Athlon XP days? Did you do it yourself? What cooler did you use ? I ask because of mounting solution.
So you suggest cutting out for example little bits of foam make them a little higher then the DIE itself and place them in the corners of the substrate? How do ensure that the foam pieces are of the same height or does it not matter that much if the are pretty soft and flexable?
Also I have seen this video : http://www.youtube.com/watch?v=AfEDateV6ik
Looks too easy but I wonder if it works on 3770K IHS due to the recessed area where the Bracket connects to.
And just clarify myself I am talking about air coolers of the weight like TRUE and upwards.

[G.Foyle]

I did it with Prolimatech Super Mega, using two flat aluminium spacers between mounting hardware and cooler base to compensate for the thickness of missing IHS. I piled several layers of tape on both sides of the die and measured the thickness with digital calipers to make sure my tape spacers are the same height as the die.

[bamtan2]

I did it with Prolimatech Super Mega, using two flat aluminium spacers between mounting hardware and cooler base to compensate for the thickness of missing IHS. I piled several layers of tape on both sides of the die and measured the thickness with digital calipers to make sure my tape spacers are the same height as the die.

you sir are a gentleman and a scholar. can we get all the ivy bridge stripping stories and pictures in one place somehow? to document this great change of scenery.

who will be first to sell ivy bridge shims?

[bamtan2]

i will run my naked 3770k on water,i believe its possible to prime it at 5GHz or more

I want to know how a vapor chamber performs when attached directly to the ivy bridge die. I'm guessing with the right noisy fan it performs just like any water cooling. come on come on!!!

Dynatron R15: 435g copper vapor chamber, 90.0 x 90.0 x 27.0 mm, rated 150W for socket 2011
http://www.dynatron-corp.com/en/prod...98&id=244&in=0

[Chukundur]

I got a drop of almost 15 degrees on my 3570K since I changed the original TIM with MX-3. Unfortunately overclocking does not change and under Prime95 x64 AVX - In-place large FFTs, it is unstable at the same frequency of 4700MHz of any voltage.
Before

After

[BMWMASTER81]

But now you can put some more vcore on it. Maybe it helps to clock 100 Mhz higher.

[Endgame]

Anyone think indigo Xtreme would work under the IHS? My concern would be repeated reflow given the temps, but on the plus side it would attach the IHS back to the core a little more solidly than paste alone. Also, any thoughts on lapping the underside of the ihs, or is it already flat?

[st0ned]

You would need an special edition of IX since I don't see any of the actual types of IX reflowing on the core surface.

[PcCI2iminal]

Naked core on water


unmodded 3770K at same clock and voltage range ,just for comparo


naked core with AS5,temps dropped 10c !!


5GHz!!!!

[ChinStrap]

Naked core on water

http://downloads.criminalcafe.com/re...898%20copy.jpg

great pic. thank you for sharing.

[Kallenator]

Wow! Brave, reminds me of the Socket-A days, bring on the safety shims for me nerves

I suppose the block's retention screws pushes the CPU hard enough down to not need the LGA metal lever fastener?

[Chukundur]

22 degrees ambient temperature. SilverArrow. Noctua NT-H1.

[itznfb]

Is anyone taking the time to lap their IHS as well?

[Defiant83]

Hi All,

I have been playing with Direct Die cooling myself and came across a few issues, I am using an EK Supreme HF to water cool the chip, however with the IHS removed and the Stock retention bracket in place the block doesn't make contact with the top of the die. I have also tried removing the retention bracket and tried to get the CPU block to make contact with the socket, however this has resulted in no post with CPU LED illuminated each time. I have when using this method tried to pack the block so that it applies more preassure to the chip but due to the block only making contact with the Die I don't think the load was even enough across the whole chip so didn't want to push it any further.

I have some small pieces of copper to hand and was going to cut these so that they pack the Die up so it is level with the Retention bracket and then mount the block on top of that. I will get some pics later if I can to show what I mean if my description isn't clear. Unless you guys have any other thoughts on what I can do to resolve this.

[itznfb]

Criminal must have been super careful when he did that
Back in the day I would buy mounting spacers that you would use in place of the bracket that would align the block properly when seating on a naked die.

[Defiant83]

Criminal must have been super careful when he did that
Back in the day I would buy mounting spacers that you would use in place of the bracket that would align the block properly when seating on a naked die.

I might have to look at cutting the old IHS so that it will spread the load once the block is mounted. Thanks for your input!

[itznfb]

I might have to look at cutting the old IHS so that it will spread the load once the block is mounted. Thanks for your input!

The easiest solution is lapping the IHS so that it has the least impact possible. If you want to run naked the next would be to get some plastic type of washers to put on the mounting screws of a water block so that they are screwed down to the right height and pressure is applied evenly.

[bamtan2]

I have some small pieces of copper to hand and was going to cut these so that they pack the Die up so it is level with the Retention bracket and then mount the block on top of that.

don't waste your time. switching from one precisely crafted piece of copper to a random hack piece of copper is not going to prove fruitful. it doesnt matter if it is a little thinner

[Astennu]

I also removed the IHS. I gained 12c when using 2x MX4 vs 1x AS5. Room temperatures where both the same 24c and i was running 4500 MHz 1.25v