Thermal Grizzly Kryosheet

AStaley · 21 Sep 2023 at 14:56

Anyone used Thermal Grizzly Kryosheet, https://www.thermal-grizzly.com/en/products/625-kryosheet-en

Reading up on it looks like it performs on par with their paste, maybe a degree or two cooler but easier to apply and won't need repasting. But looks like it's brittle, so wondering how easy it is to use with an AIO which needs to be tightened down and how does it cure with the heatsink to deal with the small imperfections that the pastes fill in.

AWitney87 · 21 Sep 2023 at 18:29

Hello AStanley.

A video was posted about a month ago reguarding your Kryosheet. It's in german, however, it does show how he mounted it with his aio and the tempreture result/ comparison.

You might need a german translater to get the finer details in the video, as I only speak english and very fluent in speaking rubbish too

.
It looks like a straight forward easy plug and play.

In all fairness to the sheet, it looks alright.

https://www.youtube.com/watch?v=fAxjjDX1qDA

.

Jamin280672 · 21 Sep 2023 at 21:13

He did brush over it at the end of his video for his direct die water block, and I was considering it, but it looks a bit big and not sure if it can be cut down or not, it's made from graphene, he provides a sticker with the direct die water block to protect the SMDs from the overhang of the sheet, you have to be careful as it is electrically conductive.

I saw another video where he was showing it off at Taipei 2023 and said it was going to cost at least $20 which in rip off Britain equates to £20 plus postage, a bit pricey if it's no good for my use, especially when I can get liquid metal ultra which performs better for £7.99.

Jump to 14.45... https://youtu.be/qKFYawQOKJo?si=9iCUhewMhnaI1CEZ

Begining of this video he discusses it https://youtu.be/v6PfZ9CJVUg?si=IhM-C29FhNwW1-6E

srekal34 · 22 Sep 2023 at 06:32

Better get PTM7950, cheaper and better performing.

AStaley · 22 Sep 2023 at 09:52

I am really tempted, just because I want to see how it performs.

Mr Evil · 22 Sep 2023 at 13:29

Hardware Unboxed tested one in the middle of their AIO-cooled 4090 review, with these results:

* Liquid metal (stock): 58°C
* Kryosheet: 60°C
* Hydronaut: 67°C

That looks really good to me, and makes me want to buy some to try on my CPUs.

AStaley · 3 Oct 2023 at 20:48

Possibly the least scientific test I’ve seen, but interesting enough; https://youtu.be/pTRk52LpRco?si=RMKONdg7ygtmagWP

PaulW · 9 Oct 2023 at 11:11

i like the idea of these not being paste. Paste to me always concerns me with the best way to apply it. Did it cover the areas it needs to effectively. sheets take that concern of mine away, you don't need to reapply it and there is no mess. only down side it wold seem is the cost and that it is conductive.

Mr Evil · 9 Oct 2023 at 16:58

I bought one. The 38 x 38mm size is exactly right for an AM4 CPU. I put it in my server, which has a passively cooled 3400GE Pro in it. I thought this might be a good use for it, since it's hard to keep a CPU cool with no fans.

The result was... exactly the same as the old thermal paste (not sure what thermal paste exactly - it was preinstalled on the heatsink). That's not bad for something that will never dry out, and should be reuseable, but definitely not worth the high price.

The CPU only uses 40W at full load though, which I suppose is not enough to make a measureable difference. I might buy another and try it on my 140W 5950X.

GuruJockStrap · 25 Oct 2023 at 11:06

Mr Evil said:
The CPU only uses 40W at full load though, which I suppose is not enough to make a measureable difference. I might buy another and try it on my 140W 5950X.

Would be interested in your results if you did.

These sheets/pads interest me as they are a pretty much set it and forget it thing. Less mess, guaranteed perfect application and no thermal paste pump out issues over time. If they give comparable thermal performance to decent pastes then even better.

Mr Evil · 19 Nov 2023 at 15:26

I purchased one more for my PC. As mentioned before the CPU is a 5950X. The power limit is at the default 142W, though HWiNFO64 records an actual max. package power of 135W. It's air-cooled.

I did the following test:

Leave the sytem idling for a few minutes to let temperatures stabilize.
Start recording CPU temp and fan speed in HWiNFO64
Run the multithreaded benchmark in Cinebench R23

This isn't a very scientific test; I didn't try to fix any variables, and just let the fan, clocks etc. run at whatever speed. Ambient temperature was 22°C. The Cinebench scores were effectively the same for both runs (0.5% difference).

Here is a chart of the results:

The temperature was almost exactly the same, but the fan speed was slightly higher with the KryoSheet, suggesting that it's very slightly worse.

the Kronaut paste had been there since 2020, and when I took the heatsink off I could see that it had dried up, so maybe fresh paste would have been better. I won't have to worry about that anymore though, since the KryoSheet won't dry out.

So, I think the conclusion is the same as last time: It performs well enough compared to paste, and has the advantage of not drying out, but it's too expensive.

Jamin280672 · 20 Nov 2023 at 21:00

Mr Evil said:
I purchased one more for my PC. As mentioned before the CPU is a 5950X. The power limit is at the default 142W, though HWiNFO64 records an actual max. package power of 135W. It's air-cooled.

I did the following test:

Leave the sytem idling for a few minutes to let temperatures stabilize.

Start recording CPU temp and fan speed in HWiNFO64

Run the multithreaded benchmark in Cinebench R23

This isn't a very scientific test; I didn't try to fix any variables, and just let the fan, clocks etc. run at whatever speed. Ambient temperature was 22°C. The Cinebench scores were effectively the same for both runs (0.5% difference).

Here is a chart of the results:

The temperature was almost exactly the same, but the fan speed was slightly higher with the KryoSheet, suggesting that it's very slightly worse.

the Kronaut paste had been there since 2020, and when I took the heatsink off I could see that it had dried up, so maybe fresh paste would have been better. I won't have to worry about that anymore though, since the KryoSheet won't dry out.

So, I think the conclusion is the same as last time: It performs well enough compared to paste, and has the advantage of not drying out, but it's too expensive.

I think derbauer was comparing Kryosheet to LM not Kyronaut, when I tested it, idles where only a degree or 2 higher with the sheet on my delidded 7900, but loads were about 7 or 8oC higher, so I took it off and went back to LM.