[EOL] Arctic MX-5 is here!!Tests incoming! Completed. Now its MX-6 testing time!

[yotano211]

Guru3D Thermal Paste Roundup - Round 2 (2021)

It has been almost two years since we made part 1 of Thermal Paste Roundup. You gave us good feedback for it; thus, we have prepared the second round of that article. As a short reminder - you should already know that the TIM (thermal insulation material) is often overlooked when setting up the...

www.guru3d.com

just posted today it looks like. looks neat and on topic lol

looks like the original kryonaut (non-extreme) is the best value overall. $10... and much better than MX-5 when on air cooler... with the AMD chip... with Intel the difference is much less... very odd...

and non-conductive. hmm.

anyone want to buy my big tube of mx-5 for like 8 bucks free ship? lol

I guess you dont like it.

 

[lexluthermiester]

Guru3D Thermal Paste Roundup - Round 2 (2021)

It has been almost two years since we made part 1 of Thermal Paste Roundup. You gave us good feedback for it; thus, we have prepared the second round of that article. As a short reminder - you should already know that the TIM (thermal insulation material) is often overlooked when setting up the...

www.guru3d.com

just posted today it looks like. looks neat and on topic lol

looks like the original kryonaut (non-extreme) is the best value overall. $10... and much better than MX-5 when on air cooler... with the AMD chip... with Intel the difference is much less... very odd...

and non-conductive. hmm.

anyone want to buy my big tube of mx-5 for like 8 bucks free ship? lol

Those results look weird. It would seem they didn't test MX-5 on the Intel side of things. Wondering why not...

 

[bobbybluz]

Those results look weird. It would seem they didn't test MX-5 on the Intel side of things. Wondering why not...

No Gelid GC Extreme tested either, usually one of the best pastes.

 

[Falkentyne]

Looking forward to the results, especially the copper anti-seize grease.

Innovation cooling seems to be selling the EYG-R pads (softer than the older EYG-S) in a 5 pack now. They say they are one time use only, while the original IC pads are reusable.
(originals: https://www.amazon.com/Innovation-Cooling-Graphite-Thermal-Pad/dp/B07CKVW18G)

Amazon.com

I tested both of the pads on a R9 290X (reference blower) and could see no difference. Both were 5C worse than Arctic MX-5 looping Valley benchmark (86C, at 32C ambient).
Both made my i9-9900k throttle at 5 ghz @ 1.280v load (VR VOUT) in Prime95 small FFT (AVX disabled). (TFX and Kryonaut Extreme and MX-5 were all within 1C of each other on a NH-D15 + 2x Silent Wings 140mm fans). I used to run my NH-D15 with 2x Noctua 3000 RPM industrial fans (full speed) which would keep Prime95 around 85C, but now all of my industrial fans are in my case and the last two I put into my Eisbaer Extreme 280 AIO which is sitting idle (because the Arctic Liquid Freezer II 360 outperforms it).

Anyway if you care about long term reliability, you can try these pads. They can't possibly be worse than GD900.

Note: you will probably save money if you bought the pads directly from digikey. They are in 0.25mm and 0.35mm variants, as I think you get a 100mm * 100mm square.

 

[Keullo-e]

Picked up a syringe as well at last. Haven't tested yet tho as I'm too lazy to uninstall the waterblock (and GPU cooler) for a possible few C drop over MX-2.

 

[95Viper]

Stay on topic.
Stop the insults and personal attacks.
Discuss the topic... not each other!
Be civil.

Thank You
And, be kind to others. Karma is known to bite back.

 

[Shrek]

On the related topic of longevity; part of the shelf life of thermal grease might be separation of the silicone oil and filler. Seems to me better to get a tub rather than syringe as one can remix with a tub.

 

[Space Lynx]

On the related topic of longevity; part of the shelf life of thermal grease might be separation of the silicone oil and filler. Seems to me better to get a tub rather than syringe as one can remix with a tub.

On same hand, I really don't need any paste other than like maybe one change every 2-3 years. I suppose it depends how much you own though yeah.

 

[lexluthermiester]

Ok folks, MX-5 is in the house once again and the testing begins this weekend! (I have to make some time.)

Sorry for the crappy image, my phone camera isn't the best..

 

[Keullo-e]

Ok folks, MX-5 is in the house once again and the testing begins this weekend! (I have to make some time.)
View attachment 201051
Sorry for the crappy image, my phone camera isn't the best..

Didn't even know that there's a version with that spatula, mine doesn't seem to have one (still unopened on its box). Those are kinda useless if you ask me.

I could finally test the paste though by repasting my HD 7970, I just need to run Superposition few times to see how it warms up with that paste it has now..

 

[lexluthermiester]

Didn't even know that there's a version with that spatula, mine doesn't seem to have one (still unopened on its box).

I didn't actively choose it, but I think I'll try it.

I just need to run Superposition few times to see how it warms up with that paste it has now..

Thinking I'm just going to use Prime95 to do the testing this weekend.

 

[Keullo-e]

Thinking I'm just going to use Prime95 to do the testing this weekend.

Too lazy to take the block/pump unit of my Eisbaer just to repaste the CPU, it has MX-2 now and it's fine.

 

[GerKNG]

i've now finished testing my 4 week long application on my 6900XT merc 319 black edition.
the paste survived the 340W powerlimit with no issues and performs identical compared to the day of application.
and it beats Noctua NT-H2 by around 3°C on the Hotspot and 2°C on the Edge temperature at the same ambient temperature and fan speed.

 

[lexluthermiester]

i now finished tested my 4 week long application on my 6900XT merc 319 black edition.
the paste survived the 340W powerlimit with no issues and performs identical compared to the day of application.
and it beats Noctua NT-H2 by around 3°C on the Hotspot and 2°C on the Edge temperature at the same ambient temperature and fan speed.

That's intense! I'm not doing anything like that, wondering if I should even bother with testing now.

What do you all say? Would everyone like me to still run the testing as planned?

 

[Athlonite]

Yes run the tests

 

[Keullo-e]

Alright, repasted and cleaned my HD 7970 Matrix (@ Platinum bios), the card is on its factory OC (1100/1650). With old TIM, 3 runs of Superposition gave maximum of 88C. Now it maxed at 85C.

Not that huge of an improvement, but drop in temps is always nice. I'll repaste my 1080 Ti later and check how it behaves.

 

[GerKNG]

Alright, repasted and cleaned my HD 7970 Matrix (@ Platinum bios), the card is on its factory OC (1100/1650). With old TIM, 3 runs of Superposition gave maximum of 88C. Now it maxed at 85C.

Not that huge of an improvement, but drop in temps is always nice. I'll repaste my 1080 Ti later and check how it behaves.

identical fixed fan speed?

 

[Keullo-e]

identical fixed fan speed?

Ah, didn't test with that and now it's too late.

 

[lexluthermiester]

Ah, didn't test with that and now it's too late.

To be fair, given the fan curve that exists on the Radeon 7970, I'll bet the improvement was a bit better as the fan likely wasn't ramping up as high.

 

[Keullo-e]

To be fair, given the fan curve that exists on the Radeon 7970, I'll bet the improvement was a bit better as the fan likely wasn't ramping up as high.

Though it's a beefy Asus ROG Matrix card (takes even three expansion slots) so I didn't hear any difference in noise level as the system is noisy even with a passive GPU.

 

[lexluthermiester]

Update: I have to delay testing MX-5 for a few days. Personal life has gotten in the way. Will get to it Tuesday or Wednesday.

 

[Nike_486DX]

plz make sure to include Prime95 SmallFFTs with AVX enabled and stuff.

 

[8BitZ80]

Update: I have to delay testing MX-5 for a few days. Personal life has gotten in the way. Will get to it Tuesday or Wednesday.

I spent a heap of money to do a big shootout with at least a dozen thermal pastes, but I abandoned that plan. Exams and a new job have drawn my attention away from it. As a general rule I have found that unless you're running expensive Thermal Grizzly or cheap-ass stuff like Arctic Alumina, everything else performs damn near identical and differentiating them is like splitting hairs. GD900 is the only exception to that trend (it performs great and costs nothing) and Arctic MX-5 is the best option for long term high performance without annual re-applications. At least with a 120W TDP processor I haven't been impressed by anything else that I've tested.

 

[Shrek]

GD900 needs annual re-application?

 

[lexluthermiester]

plz make sure to include Prime95 SmallFFTs with AVX enabled and stuff.

The system I'm going to be testing with does not have AVX(Dell T3500 with a Xeon W3680), but yes Prime95 will be used. I might even use Intel Burn Test.