Aqua Computer and Rockit Delidding Tools tested on Core i7-7700K

[Franzen4Real]

I really enjoyed this write up, thanks for sharing!

It does make one laugh though that Intel spends billions in R&D to make lower power/lower heat CPUs for devices such as ultrabooks, Surface Pro etc. and something easily done by DIY'ers has had far more impact on temps than generations of die shrinks...

 

[SPYUI]

It there any trouble to delid afterward when relid with superglue ? @VSG have you try it yet ?

 

[Aretak]

To be fair to Intel, they do. Devil's Canyon was the acknowledgement of this very thing. Doesn't mean they did everything possible though.

The problem is, the "improved" thermal compound they shouted from the rooftops about for Devil's Canyon was still crap. And it's still crap to this day, as these results prove. Intel are obviously well aware of the problem, but they've done next to nothing to address it. And why would they? Using the cheapest thermal compound they can find saves them 0.01 cents per chip and they don't guarantee any overclocking ability. It's good enough to run at stock, so that's fine in Intel's eyes. Especially when you might upgrade faster if you can't get more performance out of your current CPU due to it getting too hot when you try and overclock it...

 

[mcraygsx]

I sometimes wonder if INTEL corporation even bother to read these kind of enthusiast articles. This has been a trend for past couple of years to use a toothpaste instead of soldering. Even the likes of AMD with very low RnD can affoard to solder the chips.

 

[Dominix]

It there any trouble to delid afterward when relid with superglue ? @VSG have you try it yet ?

SuperGlue is one-way ticket bud.
u will fubar PCB, not worth to do it, better to use silicone so u can delid later, just like I did for first time and had to delid it again, ez pz

 

[GorbazTheDragon]

I delidded my 4790k a few years back, has a small scratch on the pcb surface from where i got a bit impatient handling the razor blade... It still runs pretty warm, up in the 80s under artificial load with an NH-U14S (4.7GHz@1.26v) so I haven't really pushed the voltages up even more, it's not the greatest of silicon as it needs 1.33 to hit 4.8 and 1.38 to get 4.9. I used to get it to run at 4.8@1.28 but it seems to have deteriorated from running that (was a different cooler ran at 85c+ gaming load).

Nice review, but I cant help thinking that this should not be needed on such a cpu. More than 20 degrees improvement should make Intel reconsider their thermal solution. Considering that buyers lose there guarantee by doing this.

As others have mentioned, people were getting 20s regularly on ivy delids. I doubt intel will change their ways, maybe with the ryzen pressure, but I still doubt it.

 

[Shoggy]

In practice, however, there is a small amount of give on all sides to where you can move the centering tool laterally and along with it the IHS.

And the simple reason for that is that the centering tool has been used in the wrong orientation. The cut-outs must point to the left and right side as also indicated in the manual. Then it will enclose the heatspreader with almost no play

 

[VSG]

SuperGlue is one-way ticket bud.
u will fubar PCB, not worth to do it, better to use silicone so u can delid later, just like I did for first time and had to delid it again, ez pz

Not necessarily. Have it in the corners, use dilute acetone and it can come off. I have done it, and the Rockit guy has also multiple times.

 

[Upgrayedd]

Was there any thermal material applied to the underside of the IHS? I seen where it was applied to the die. Im curious because I have seen others do this and I thought that they used too much material.

 

[VSG]

And the simple reason for that is that the centering tool has been used in the wrong orientation. The cut-outs must point to the left and right side as also indicated in the manual. Then it will enclose the heatspreader with almost no play

Now you tell me! I will blame it on the manual then

But seriously though, thanks for letting me know. I will mention it in the article.

 

[Shoggy]

Now you tell me! I will blame it on the manual then

But seriously though, thanks for letting me know. I will mention it in the article.

Yeah... We also feel a bit stupid that no one has realized that at first when reading the review. We were only wondering why you had problems there and after a closer look at the photos I saw what is going on

 

[VSG]

Was there any thermal material applied to the underside of the IHS? I seen where it was applied to the die. Im curious because I have seen others do this and I thought that they used too much material.

I did apply a thin coat of the liquid metal. I am not sure I would do the same with non conductive, dense TIM however.

 

[MrGenius]

Where's @cadaveca and the anti-delidding crew? With all the at-length explanations as to why you should NEVER do this and how RIGHT Intel is to use 3.1 W/mK TIM under the IHS.

 

[The Von Matrices]

I always thought you needed to cut the glue with a razor before delidding, but these tools seem to not do that. Is there an advantage to cutting the glue first?

Also, one thing that no one is considering about Intel's stock TIM is that it has to last the lifetime of the processor, which can be a decade or more. The replacement TIMs with better performance aren't guaranteed to work for more than a few years. Yes, reapplying TIM is easy for us, but most people aren't going to ever touch the insides of their PCs.

 

[Upgrayedd]

I did apply a thin coat of the liquid metal. I am not sure I would do the same with non conductive, dense TIM however.

Maybe you could clear this up for me. Do you think it matters with liquid metal to put it on the underneath of the IHS if you have the die completely coated already? I know that by scraping the old silcone you get closer to the die with the IHS. This makes me think that one should put TIM on the underside of CPU cooling blocks if it goes on the underside of IHS as well. I just always thought you wanted to get as little as you could between the cpu and cooling solution. The way I think I would do it is putting liquid metal TIM on the die, reseat IHS as lowas possible, then apply the same TIM to the top of the IHS. I could be way wrong too, I just like to understand a little better for when the day comes my 4790K needs an overclock.

 

[fullinfusion]

Where's @cadaveca and the anti-delidding crew? With all the at-length explanations as to why you should NEVER do this and how RIGHT Intel is to use 3.1 W/mK TIM under the IHS.

LMAO I was just thinking the same thing

Nothing like a razor blade and a water block to do the trick

Maybe you could clear this up for me. Do you think it matters with liquid metal to put it on the underneath of the IHS if you have the die completely coated already? I know that by scraping the old silcone you get closer to the die with the IHS. This makes me think that one should put TIM on the underside of CPU cooling blocks if it goes on the underside of IHS as well. I just always thought you wanted to get as little as you could between the cpu and cooling solution. The way I think I would do it is putting liquid metal TIM on the die, reseat IHS as lowas possible, then apply the same TIM to the top of the IHS. I could be way wrong too, I just like to understand a little better for when the day comes my 4790K needs an overclock.

Yes it makes a huge difference, you WANT to use the liquid metal under the IHS and also on the die....

I did 3 delids and when the liquid metal was just used on the die itself it didn't fully take to the underside of the IHS.

see what I mean..

Delid 7700K and applying CLU + results

 

[VSG]

Maybe you could clear this up for me. Do you think it matters with liquid metal to put it on the underneath of the IHS if you have the die completely coated already? I know that by scraping the old silcone you get closer to the die with the IHS. This makes me think that one should put TIM on the underside of CPU cooling blocks if it goes on the underside of IHS as well. I just always thought you wanted to get as little as you could between the cpu and cooling solution. The way I think I would do it is putting liquid metal TIM on the die, reseat IHS as lowas possible, then apply the same TIM to the top of the IHS. I could be way wrong too, I just like to understand a little better for when the day comes my 4790K needs an overclock.

With the use of super glue, the IHS was close enough to where there was good contact of the liquid metal on the underside as well. With silicone glue, go ahead and apply some on the IHS underside separately as you may not have enough contact otherwise. Easiest way to tell is to de-lid again irrespective of the glue used and see if the IHS had a good rectangular TIM spread on it also. If you are not comfortable de-lidding after having used super glue then add a thin layer on the underside since a thicker layer of liquid metal is still going to be way better than not enough contact and an air layer in between.

 

[Fluffmeister]

An elegant well designed product... that ultimately shouldn't even exist.

Great review @VSG

 

[fullinfusion]

With the use of super glue, the IHS was close enough to where there was good contact of the liquid metal on the underside as well. With silicone glue, go ahead and apply some on the IHS underside separately as you may not have enough contact otherwise. Easiest way to tell is to de-lid again irrespective of the glue used and see if the IHS had a good rectangular TIM spread on it also. If you are not comfortable de-lidding after having used super glue then add a thin layer on the underside since a thicker layer of liquid metal is still going to be way better than not enough contact and an air layer in between.

what chip you talking about?

 

[VSG]

what chip you talking about?

The i7-7700K. I saw your post and not sure why CLU broke apart like that. Either way, I agree that it is best to have it applied on the die and IHS underside.

 

[Upgrayedd]

With the use of super glue, the IHS was close enough to where there was good contact of the liquid metal on the underside as well. With silicone glue, go ahead and apply some on the IHS underside separately as you may not have enough contact otherwise. Easiest way to tell is to de-lid again irrespective of the glue used and see if the IHS had a good rectangular TIM spread on it also. If you are not comfortable de-lidding after having used super glue.

What's up with the super glue after delid? Is it much harder to delid or something? Couldn't you bake it in the oven for a bit to warm the glue? I feel like I should have made a forum post but its on topic.

 

[Solidstate89]

https://siliconlottery.com/products/delid

Sometime's he runs a delidding sale, got my 6700K delidded for $29.99.

But back then, January 2016, we did not have these amazing DIY delidding tools/options presented in this review. l

Cool. Never heard of that site. I'll have to keep that in mind (depending on my CPU choice) when I upgrade my desktop.

 

[fullinfusion]

The i7-7700K. I saw your post and not sure why CLU broke apart like that. Either way, I agree that it is best to have it applied on the die and IHS underside.

it was because I didn't apply the metal to the underside of the lid.. its funny stuff to work with, and why its a bit challenging to get it stick, but once it sticks its easily spread

 

[hapkiman]

Great review. Since these products for delidding are now so readily available, and it obviously DOES benefit temperatures, a simple solution would be for Intel to allow delidding without voiding the warranty.

Just as long as you don't physical crack, bend, or obviously damage the CPU-PCB or the IHS, then you should be covered. There could be some kind of indicator or internal seal that breaks when to delid so a buyer of a second hand proc would know if he was buying a delidded chip.

I'm just saying if they want to use a cheaper thermal material, they have to expect this type of response, and companies developing such products.

Last proc I delidded was with a straight edge razor blade. They didn't even have these things.

 

[Logun]

I purchased the Rockit kit in anticipation of doing my first ever delid on my 7700k. Are the any more tips and tricks? I was not going to re-lid it as I was going to let the pressure from the cooler handle that.

My worry is in spreading the liquid metal getting some on the PCB when the pressure is applied to the IHS when strapping it in and adding the cooler - so I'm not sure how I would apply TIM to the underside of the IHS without risk of getting it over the PCB for not having it lined up 100%?