Delidding Intels 7700k drops temps by almost 30C

[Lays]

3 minutes ago, Dabombinable said:

what's the difference going to be though? The silver based compound is also liquid metal.

This is what I plan on using:

https://www.jaycar.com.au/heatsink-compound-25-silver-1-5g/p/NM2013

https://www.jaycar.com.au/medias/sys_master/images/h27/h5e/8886994862110/NM2013-msdsMain.pdf

trust me bro just get CLU, i've been doing this for a long time, used boatloads of paste and CLU kicks everythings ass WHEN BETWEEN THE IHS AND DIE

 

 

[MageTank]

3 minutes ago, Dabombinable said:

what's the difference going to be though? The silver based compound is also liquid metal.

This is what I plan on using:

https://www.jaycar.com.au/heatsink-compound-25-silver-1-5g/p/NM2013

https://www.jaycar.com.au/medias/sys_master/images/h27/h5e/8886994862110/NM2013-msdsMain.pdf

I cannot for the life of me, get those pages to load. That being said, I've used CLU, and can vouch for it's effectiveness. Your product (from what google tells me) contains far more silicone than metal, meaning it's effectiveness will almost certainly be less than that of CLU. If you are going to be putting in the effort of delidding, i'd say go for broke, and do it in the best way possible.

 

Buy some clearcoat nail polish, cover the gold contacts near the bare die for added safety, and buckle down and grab some CLU. Cut the brush bristles in half, and lightly dip it in the CLU syringe, and using only that amount, thinly paint the die. Repeat for the underside of the IHS (the side that will be making contact with the die) so that the CLU adheres to itself when IHS is applied. This results in the best contact, and will give you the results we have.

[Lays]

1 minute ago, MageTank said:

I cannot for the life of me, get those pages to load. That being said, I've used CLU, and can vouch for it's effectiveness. Your product (from what google tells me) contains far more silicone than metal, meaning it's effectiveness will almost certainly be less than that of CLU. If you are going to be putting in the effort of delidding, i'd say go for broke, and do it in the best way possible.

 

Buy some clearcoat nail polish, cover the gold contacts near the bare die for added safety, and buckle down and grab some CLU. Cut the brush bristles in half, and lightly dip it in the CLU syringe, and using only that amount, thinly paint the die. Repeat for the underside of the IHS (the side that will be making contact with the die) so that the CLU adheres to itself when IHS is applied. This results in the best contact, and will give you the results we have.

I didn't even bother painting the underside of IHS, just scraped some of the black glue shit off, painted the die with CLU, plopped the IHS on and bam

[MageTank]

Just now, Lays said:

I didn't even bother painting the underside of IHS, just scraped some of the black glue shit off, painted the die with CLU, plopped the IHS on and bam

You madman!

[Lays]

1 minute ago, MageTank said:

You madman!

[MageTank]

8 minutes ago, Lays said:

-snip-

[patrickjp93]

20 minutes ago, MageTank said:

Don't use aluminum, problem solved. Copper is superior anyways, unless weight is important to you. That being said, you can get away with nickel plated aluminum (if any heatsinks exist) as long as you don't accidentally scratch through the plating. 

 

EDIT: As @Lays said, for the bare die, it's perfectly fine. If using it on top (don't really recommend, as it doesn't do much) follow my advice above. Though, I personally just use gelid on top, lol. 

PFFT, no EK Indigo? 

[Bouzoo]

To the both of you, noz to quote you on phone, how would CLU work on a laptop when compared to NT-H1? 

[patrickjp93]

1 minute ago, Bouzoo said:

To the both of you, noz to quote you on phone, how would CLU work on a laptop when compared to NT-H1? 

Assuming a non-aluminum heatsink, very very well, but you'll have to reapply every few months since it's a more open-air environment.

[Bouzoo]

8 minutes ago, patrickjp93 said:

Assuming a non-aluminum heatsink, very very well, but you'll have to reapply every few months since it's a more open-air environment.

I plan on using it on my Asus N550JK on both CPU and GPU, didn't have the time to change the paste so if it might help... I'm curious as to what I might expect on using CLU and Noctua, or better to say if I can expect much lower temps on CLU, and if I can live with reapplying CLU once a year? @MageTank @Lays 

 

And I'm on laptop so can do proper posts.

[Lays]

13 minutes ago, MageTank said:

kek

[MageTank]

12 minutes ago, Bouzoo said:

To the both of you, noz to quote you on phone, how would CLU work on a laptop when compared to NT-H1? 

I personally had both good and bad experience with CLU. At first, it was amazing, as it dropped temps by about 15C on my brothers XPS13. Then a couple weeks ago, disaster struck, and temps soared high. 

 

 

It completely lost it's liquid state, and turned extremely coarse. In fact, after using a microfiber cloth and some 91% isopropyl, this was as clean as I could get the die:

 

 

In order to get the heatsink clean, I had to use sandpaper. It was indeed a copper heatisnk, so it's not as if the gallium caused corrosion. I found a few others on OCN complaining about the same effect, but I'll need to do more tests to confirm. It might be chalked up as old CLU, as this tube was about 6 months old, and not stored in the best conditions. 

 

EDIT: I've used CLU on this 6700k for almost 8 months now, and temps are relatively the same as the day I first applied. Most people on OCN recommend reapplying once a year, to prevent the staining effect above, but as far as thermal performance goes, I have not seen drastic changes for prolonged usage. Maybe 1-3C at most. 

[Bouzoo]

7 hours ago, MageTank said:

Snip

I did a bit of research and as it seems to me, the delta between CLU and Noctua shouldn't be that big, few C at most. I still think imma go with Noctua, but I will keep CLU in mind. And if that what happened to your brothers XPS were to happen to me, it might be a problem since I rarely have time to open my laptop and deal with it. 

[Tieox]

If you de-lid your CPU with a chainsaw, it's temperatures are even better*

 

 

 

 

 

 

 

 

 

 

 

 

 

*note method is still in Alpha.

[Syntaxvgm]

On 12/15/2016 at 9:31 PM, Ryan_Vickers said:

Please update this to fit the news posting guidelines.

 

---

 

That's huge.  I had heard it was running hotter than Skylake but if this is why, then there's hope - maybe it's not a bad chip after all, just a terrible thermal solution.  Hopefully when they ship them they will have fixed this by essentially factory-deliding them  

intel are the masters of making things worse when it comes to cpu lids and the compounds they use in-between them. 1000 dollar CPUs and they cheap out on the fucking lid and paste. 

[Syntaxvgm]

On 12/16/2016 at 5:09 AM, Gibbons said:

https://forums.anandtech.com/threads/my-i7-7700k-has-arrived-insights-benchies-overclocks-inside-now-with-delid.2493250/

 

This is the real source. Scroll down and see his previous testing.

 

He used a big water cooler(corsair h110i) and got avg 83 degrees with 4.7 Ghz @ 1.264v.

 

After delid and using another big water cooler (Kraken x62) he got avg 72 degrees with 5.1Ghz - 1.376v.

 

On 12/16/2016 at 5:14 AM, Master Disaster said:

Jesus, don't these people understand how variables work?

 

For a consistent result they must keep all variables the same across all tests. This moron changed just about every single thing he could, different cooler? Check. Different clock speed? Check. Different voltage? Check.

 

He might as well say green is better than purple because my bacon tastes of sunshine.

yah, that's stupid. still a pretty big difference though, reminds me of the 4770k. 

[Dabombinable]

59 minutes ago, Syntaxvgm said:

 

yah, that's stupid. still a pretty big difference though, reminds me of the 4770k. 

With the amount people pay for the K series, it should always come with the best possible TIM.

[ace_cheaply]

4 hours ago, Syntaxvgm said:

intel are the masters of making things worse when it comes to cpu lids and the compounds they use in-between them. 1000 dollar CPUs and they cheap out on the fucking lid and paste. 

None of the $1000 cpus have cheap paste.  The entire e series is soldered.

[Dabombinable]

4 hours ago, ace_cheaply said:

None of the $1000 cpus have cheap paste.  The entire e series is soldered.

And that's only possible due to the size of the die.

[Mitchell 3]

On 12/15/2016 at 9:42 PM, brighttail said:

Even if the temperatures are 1/2 what is claimed, it could be note worthy.  Obviously more reviews are needed but this is the second or third article I have seen about their i7700K chips running hotter than the previous generation. 

lol it's all that hevc 10-bit decoding ....

[ace_cheaply]

1 hour ago, Dabombinable said:

And that's only possible due to the size of the die.

Well, the size of the die doesn't excuse the quality of the thermal compound, but as far as soldering goes that's true.

[S w a t s o n]

As for people talking about TIM. I recommend MX-4. Why?  Well take a look. I use it on the die and the heatsink and for my GPU. I actually tried Indigo Extreme (on the IHS obv) and it was honestly garbage, CLU has it's issues and MX-4 is pretty much on par.

Spoiler

 

 

[Dabombinable]

22 minutes ago, ace_cheaply said:

Well, the size of the die doesn't excuse the quality of the thermal compound, but as far as soldering goes that's true.

I know-Intel should use the best with the price of the K series.

[jagdtigger]

12 minutes ago, Dabombinable said:

I know-Intel should use the best with the price of the K series.

But they do not do that, the stuff that was under the IHS was more like "gunk" than TIM... :

Spoiler

 

[ThinkWithPortals]

This article is total bullshit.

 

- They changed the cooler from a Thermalright 6 to a Corsair H110i to an NZXT Kraken X62

- They changed the voltage from 1.34 to 1.264 to 1.344

- They changed the clockspeed from 5GHz to 4.7GHz to 5GHz again

 

There's no way they could possibly draw any kind of conclusion from such a BS test where basically every variable changed, let alone a conclusion where temps decreased by 30 degrees. They might as well have used a different CPU altogether.