Damaged 5900X & AiO plating after TG Kryonaut

[illli]

IMO the gains beyond about 600 grit wet-and-dry sandpaper are negligible. Sure, it's gonna look shiny but that has no impact on the thickness of the thermal paste layer at that point.
Just make sure the surface you are using as a base for the sandpaper is genuinely flat, I'd suggest a piece of glass sheet, stone cutting block, or a purpose made acrylic block. If you want to be sure you're lapping evenly, use a sharpie marker to colour in the IHS or coldplate and you know you're done when all the sharpie is gone, and can see if you're applying pressure unevenly to one side so that you can adjust your grip.

yep, definitely get diminishing returns, but it is really fun seeing the results, maybe not practical results, more like "how shiny can I make this"sort of way hah

 

[AsRock]

It's the best non-conductive paste on the market: https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,12.html

I've personally compared it to Noctua and Gelid solutions and it works very very well. It doesn't dry up as quickly as those pastes as well.

None of this excuses the issues encountered by the OP. Thermal Grizzly should do a better job supporting their products.

It's worthless if they have random bad batches that do this, like who the hell except people making money talking about it want to go though this BS.

 

[Frick]

So is the expectation the OP should be given a replacement CPU and cooler? Is that even the right approach? I wouldn't classify it as "hardware damage" as it's really only superficial. As for degrading thermal performance due to the scratches (or whatever they are) I don't believe that would happen. Bridging those gaps is literally the point of TIMs. Or if the expectation is some kind of compensation, what would that be? The worth of the cost of the tube?

 

[moproblems99]

Best is a matter of opinion. Kryonaut is very good, but even NT-H1 has been tested to perform better and prolimatech nano is the best non liquid metal paste.

View attachment 193860

Why is anyone arguing about margin of error differences?

 

[moproblems99]

So when another paste trumps your precious kryonaut it's "margin of error"? Psssh. Fanboy.

Sorry, I use NTH2, because that came with my cooler.

Also, there is a .26C difference between the two. I doubt the accuracy of any tech reviewer that they can consistently be that precise in application that they could repeat those temperature tests in any manner of consistency. Buy a thermal paste and use it. Unless you are going for world records, it makes no damn difference.

And put away the fanboy reference. It makes your muffintop show.

 

[freeagent]

Just throwing this out there but.. I was using Thermalright TFX, supposed to be some of the best stuff on the market.. rated slightly higher then Kryonaut. I went to reapply after swapping coolers but the line got thin less than halfway down.. Had to go to my local shop, and all they had was AS5. I am seeing 3c difference at best, and almost everyone smacks the shit out of AS5 pretty much at any given opportunity.

So.. either I am doing something right, or someone is doing something wrong. Reviews are to be taken with a grain of salt imo. You could get better results then the reviewer, or not as good..

 

[Zach_01]

Why is anyone arguing about margin of error differences?

The best and worst TIM can differ temp by 4~5C? Lets call it a margin of error.
The best and worst cooler can differ temp by 4~5C? Lets call it also a margin of error.
The best and worst case air flow can differ temp by x~xC?
The lowest and highest ambient(room) temp can....

You can see my point I'm sure. A few margins of error can result to a significant different temp and when you end up with at least 10C of difference you can't call that a margin of error anymore

Just throwing this out there but.. I was using Thermalright TFX, supposed to be some of the best stuff on the market.. rated slightly higher then Kryonaut. I went to reapply after swapping coolers but the line got thin less than halfway down.. Had to go to my local shop, and all they had was AS5. I am seeing 3c difference at best, and almost everyone smacks the shit out of AS5 pretty much at any given opportunity.

So.. either I am doing something right, or someone is doing something wrong. Reviews are to be taken with a grain of salt imo. You could get better results then the reviewer, or not as good..

I was using AS5 for many years and I consider it one of the best TIMs. Because of the very high viscosity of AS5 the trick was to apply a very thin layer and let it cure and temp cycle (warm, cold, warm, cold...) quiet some time. I was using a small plastic bag and 1 finger to apply it to both CPU IHS and cooler's cold plate by just touching the whole surface. Applying a bulb in the middle or drawing Xs didnt do any good. When I buld the Ryzen system I wasnt very satisfied by it though so I switched to Conductonaut. The ~90W R5 3600 is exactly half the wattage of my previous CPU (FX) and yet was operating at same (max) temp (65~66C). Same AS5, AIO cooler and CPU load. Because almost 80W come from a very small (about 1/4) and off center surface compared to the FX. With LM temp dropped to 61~62C.

 

[John Naylor]

1. When it 1st came out Kryonaut was twice the price as everybody else. These days no more than any market leaders.

2. The was a bad batch of TIM. The batch numbers have been published and buyers are advised to check w/ TG if they are unsure whether or not they are affected.


Kryonaut - $6.99 - $8.99
Gelid Extreme- $6.99 - $8.99
Shin Etsu G751 - $3.99 - $7.50
NH-1 - $7.95 - $10.95
Kingpin - $10.99 - $12.99

I'd include AS5 in there but due to he capacitance and curing issues identified on their home page. The risk may be small for experienced users but can't justify the selection when the competition is free of these issues.

3. The problem with testing is performance changes by processor, shape, cooler, application methods, room temperature, time exposed to air and other factors. Look at 10 reviews and you will see far from 100% agreement. Heck, look at 1 review and see how performance varries by processor and cooler ... picking 4 perennial favorites.

8700k (4.8 GHz - Air) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,9.html
(1) Kryonaut / (3) Prolimatech / (4) Gelid / AS5 (16)

9900k (5.0 GHz - Air) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,10.html

Guru3D Thermal Paste Roundup 2019

Thermal paste is an often-overlooked part of most computer setups. You can simply use the pre-applied stuff, or the thermal paste that comes with your motherboard, and still get reasonable processor t... Air Cooling for i9 9900K@5.0 GHz 1.35V

www.guru3d.com

(4) Kryonaut / (6) Prolimatech / (3) Gelid / AS5 (17)

8700k (4.8 GHz - AIO) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,11.html
(2) Kryonaut / (5) Prolimatech / (3) Gelid / AS5 (20)

9900k (5.0) GHz - AIO) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,12.html
(2) Kryonaut / (5) Prolimatech / (6) Gelid / AS5 (17)

Average Ranking in 4 tests:

Kryonaut ~ 2.25 ... but ranged from 1st to 4th place
Gelid ~ 4.00 ... but ranged from 3rd to 6th place
Prolimatech ~ 4,75 ... but ranged from 3rd to 6th place
AS5 ~ 17.5 ... but ranged from 16th - 17th place


4. Ya have to chuckle at the mind set that says 4-5C difference doesn't matter, when folks will spend three times what they'd pay to get a $50 air cooler to get an AIO that is 2.5 times as loud and gains 1C. If 4-5C doesn't matter then why spend an extra $100 extra for an AIO ?

 

[thesmokingman]

Nice PK3 still kicking some serious arse.

 

[moproblems99]

The best and worst TIM can differ temp by 4~5C? Lets call it a margin of error.
The best and worst cooler can differ temp by 4~5C? Lets call it also a margin of error.
The best and worst case air flow can differ temp by x~xC?
The lowest and highest ambient(room) temp can....

You can see my point I'm sure. A few margins of error can result to a significant different temp and when you end up with at least 10C of difference you can't call that a margin of error anymore

We aren't talking about the best and worst. We are talking about the difference between 3rd and 7th. It's about 1.25C. Again, the inconsistent nature of applying paste just adds to the issues. All the top brands are close enough that you will fudge the application up before the brand matters.

 

[xdk]

@OP,
Looks like it got scratched when you twisted off the the water block from the CPU. It's surface really, its not pitted. If it was so wide spread why didn't you not buy it lolz? You came here to the Wailing wall, now go away see you in 2 weeks when you fu@k up something else.

Twisted? The ihs and cooper plate are rather smooth AF, so if even, this would show how heavy abrassive properties that thermal grease has, but its not polishing paste right?
Srsly you not see the pits, simillar to that on other threads? Then you must be blind, sorry.
I realized its wide spread after that happen, u must Google it to find something..
Before I had 3900X, used Kryo also, and demounted cooler / changed paste many times, its surface was still close to brand new when I sold it.

So is the expectation the OP should be given a replacement CPU and cooler? Is that even the right approach? I wouldn't classify it as "hardware damage" as it's really only superficial. As for degrading thermal performance due to the scratches (or whatever they are) I don't believe that would happen. Bridging those gaps is literally the point of TIMs. Or if the expectation is some kind of compensation, what would that be? The worth of the cost of the tube?

Why not? It seems obvious in case damage is deep and serious, and it simply wouldn't have happened if I had used a different grease.
In simillar thread I've linked before, OP for far less damaged Noctua cooler coldpate got compensation c.a 150% of brand new cooler price.

I got reponse from support, so will see.

 

[oobymach]

Kryonaut - $6.99 - $8.99

Not in Canada, here it's $30 a gram.

Thermal Grizzly Kryonaut Thermal Grease Paste - 1.0 Gram - Newegg.com

Buy Thermal Grizzly Kryonaut Thermal Grease Paste - 1.0 Gram with fast shipping and top-rated customer service. Once you know, you Newegg!

www.newegg.ca

 

[anachron]

@oobymach : Isn't it half this price on amazon.ca?

Amazon.ca

It's around 11€/g here on amazon.fr

 

[pantherx12]

This is why I just use GD 900.

 

[AsRock]

1. When it 1st came out Kryonaut was twice the price as everybody else. These days no more than any market leaders.

2. The was a bad batch of TIM. The batch numbers have been published and buyers are advised to check w/ TG if they are unsure whether or not they are affected.


Kryonaut - $6.99 - $8.99
Gelid Extreme- $6.99 - $8.99
Shin Etsu G751 - $3.99 - $7.50
NH-1 - $7.95 - $10.95
Kingpin - $10.99 - $12.99

I'd include AS5 in there but due to he capacitance and curing issues identified on their home page. The risk may be small for experienced users but can't justify the selection when the competition is free of these issues.

3. The problem with testing is performance changes by processor, shape, cooler, application methods, room temperature, time exposed to air and other factors. Look at 10 reviews and you will see far from 100% agreement. Heck, look at 1 review and see how performance varries by processor and cooler ... picking 4 perennial favorites.

8700k (4.8 GHz - Air) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,9.html
(1) Kryonaut / (3) Prolimatech / (4) Gelid / AS5 (16)

9900k (5.0 GHz - Air) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,10.html

Guru3D Thermal Paste Roundup 2019

Thermal paste is an often-overlooked part of most computer setups. You can simply use the pre-applied stuff, or the thermal paste that comes with your motherboard, and still get reasonable processor t... Air Cooling for i9 9900K@5.0 GHz 1.35V

www.guru3d.com

(4) Kryonaut / (6) Prolimatech / (3) Gelid / AS5 (17)

8700k (4.8 GHz - AIO) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,11.html
(2) Kryonaut / (5) Prolimatech / (3) Gelid / AS5 (20)

9900k (5.0) GHz - AIO) - https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,12.html
(2) Kryonaut / (5) Prolimatech / (6) Gelid / AS5 (17)

Average Ranking in 4 tests:

Kryonaut ~ 2.25 ... but ranged from 1st to 4th place
Gelid ~ 4.00 ... but ranged from 3rd to 6th place
Prolimatech ~ 4,75 ... but ranged from 3rd to 6th place
AS5 ~ 17.5 ... but ranged from 16th - 17th place


4. Ya have to chuckle at the mind set that says 4-5C difference doesn't matter, when folks will spend three times what they'd pay to get a $50 air cooler to get an AIO that is 2.5 times as loud and gains 1C. If 4-5C doesn't matter then why spend an extra $100 extra for an AIO ?

A quick look on Amazon


Kryonaut 8g $19.00 shipping included ( only 4 left hehe). Other wise $9 per gram + shipping.

MX-4 8g $11.99 free shipping or you can buy from amazon and pay a extra $5.

 

[R0H1T]

The best and worst cooler can differ temp by 4~5C? Lets call it also a margin of error.

That's not true & there's a good number of reasons, including design & surface area, for that. What're you gonna do about that viscous liquid/paste that can only be applied in one way?

 

[evernessince]

Best is a matter of opinion. Kryonaut is very good, but even NT-H1 has been tested to perform better and prolimatech nano is the best non liquid metal paste.

View attachment 193860

You are taking a single example and extrapolating it to the whole (and from Tom's Hardware no less, which is trash). For some god forsaken reason they decided to use an expensive semi-custom loop on a stock Intel processor as well. It's not a realistic configuration for an Intel CPU, where you spend all this extra money on a cooler and on the K variant and then don't OC nor does it properly display the difference between the pastes. The chart you linked is absolutely worthless for extrapolating any useful data from. Guru3D clearly thought their benchmark out more as they overclock the processor to see how the pastes handle the increased thermal load.

The link I posted before shows the NT-H1 quite a bit behind TG:K in not just high pressure mounting situations but air high / low and water high / low pressure mounts.. I've compared NT-H1 to Thermal Grizzly on the same system and can confirm Guru3ds results with a high mount pressure air cooler. Thermal Grizzly is better.

It's less a matter of opinion and more about looking at the whole picture and not cherry picking a single result to suite a narrative.

Is it though? Is 2°C within margin of error and then there is the obvious potential to of damaged components. Yet people obsess over that 2°C like it is going to make them an extreme overclocking god.

Non conductive pastes can't damage components. It has less to do with making anyone an OC god and more getting the most out of your PC. A $13 investment is nothing compared to the total price of the PC.

It's worthless if they have random bad batches that do this, like who the hell except people making money talking about it want to go though this BS.

That would be singular, bad batch. You'd be hard pressed to find a single company that hasn't made a mistake at some point. I agree this could certainly be handled much better

 

[ratirt]

A TG paste did that? On the cooler plate it looks like somebody rub it with a metal brush. Same on the CPU IHS when the paste was taken off it got scratched when using sand paper. I'm really skeptic a thermal paste (just by applying it) did that.

 

[Chomiq]

To the point, the expensive, "premium", "german enginereed" TG Kryonaut thermal grease scratched and damaged my expensive brand new AMD Ryzen 5900X IHS plate and copper plate of my Arctic Liquid CPU cooler. Probably well-known "bad batch" or something. I've used Kryo in the past without any issues, but It happend.

So you used it in the past with no problems and now that you run into a bad batch you'll never use it again. Ok.

This is permament hardware damage and possible heat/contact issues in the future

No it's not, as mentioned by others thermal compound is designed to fill in those gaps/pits whatever you call them (I doubt they're super deep, you could probably remove them with few swipes of fine sandpaper).

So what do you expect from TG? New CPU and new cooler? Refund for Kryo?

 

[moproblems99]

So you not only don't like the fact that another paste beat kryonaut you're such a die hard fanboy of it you can't accept that it's not the be all end all of pastes. I

Thread ban incoming but grow the fuck up buddy.

 

[damric]

The damage almost reminds me of that IC Diamond clay stuff that we all got for free on the forum a few years ago. That stuff was rough.

 

[newtekie1]

The best and worst TIM can differ temp by 4~5C? Lets call it a margin of error.
The best and worst cooler can differ temp by 4~5C? Lets call it also a margin of error.
The best and worst case air flow can differ temp by x~xC?
The lowest and highest ambient(room) temp can....

You can see my point I'm sure. A few margins of error can result to a significant different temp and when you end up with at least 10C of difference you can't call that a margin of error anymore

We aren't talking 4-5°C, we are talking 2°C. That is margin of error. No one seems to notice that every website that does thermal paste tests ever agree with each other on which the best performing paste is. Guru3d has Kryonaut outperforming NT-H1, then Toms had NT-H1 outperforming Kryonaut. If you look hard enough you can probably find a website that puts any of the popular pastes at the top. Why? Margin of error and the reality that all the top pastes perform extremely similar to each other. So unless you are willing to switch to liquid metal, there isn't really anything to be gained by stressing over which thermal paste to use. Hell, the difference between Kryonaut and Arctic Alumina is less than 3°C according to Tom's. Don't waste money on a tiny tube of paste thinking it will be any better than the cheaper huge tupe of MX4.

On the other hand, you should take into consideration other factors. Ease of use is a big one. I like a paste that I can just put a dot on the processor and stick the heatsink on. I don't consider a paste that needs to be spread by me a good paste.

But the biggest consideration is potential to damage components, like with Kryonaut and IC Diamond. I'm not going to use either one of those because they are known to damage components. I'm also not going to use liquid metal, with the exception of with delidding. And I won't use any product from IC Diamond simply because of how they handled the situation when people found out their paste was damaging components.

 

[neatfeatguy]

I just stick with AS5 - works good. Hasn't caused me any issues. Cleaning up is fairly simple and applying it is easy. It does the job and if the application of AS5 isn't working on helping keeping components cool, then something else if very wrong (not enough AS5 used or too much used, heatsink not seated properly, poor air flow, improper heatsink used that can't dissipate heat well enough for the application you're using it for).

If 1-2C is enough of a temperature difference that it makes or breaks your build....then clearly something else is wrong with your setup.

Stick with what works.

 

[Frick]

The damage almost reminds me of that IC Diamond clay stuff that we all got for free on the forum a few years ago. That stuff was rough.

So the OP has doctored the images??!!

(for the reference, find the ICD thread, it's sad)

 

[basco]

interesting: i understand damric´s post as like : he feels sad about what happend then with icd and now with kryo.