Question Thermal paste gone wrong

[hardware_bxl]

The cooler is Noctua NH-D15 Chromax, the CPU is AMD 7900x : after I removed the cooler and checked the thermal paste (see the link for photos), I noticed the thermal paste was all to one side of the cooler's contact plate, I guess this is not good. Should I change the offset for this cooler and how would I do this? or is this normal?
It's my first build, so not sure.
Also, I know it's not critically important, but the thermal paste does not look like it covers the cpu properly, now that I will re-do the thermal paste, how to do this better?
The thermal paste I used is Thermal Grizzly Kryonaut as I was told this was a good choice, but I also have thermal paste that came with the Noctua cooler.
Which thermal paste to use and how to apply it best?

Thank you all so much for your help! First time builder's appreciation is huge ;-)

Here are the images

 

[Misgar]

The cooler base plate does appear to somewhat "off center" but perhaps that's because it would also cover an Intel LGA1700 or similar rectangular CPU. I'm not sure if you can change the offset, but I think Noctua supply offset kits for some coolers.

The cooler seems to be making contact with all of the Ryzen's IHS, so that's OK, but I think you're using far too much thermal paste. A blob the size of a small pea in the center of the Ryzen is all that's required.

I use the paste that comes with my Noctua coolers, but you could wipe that off and use the Thermal Grizzly instead. I doubt you'd see much difference.

 

[hardware_bxl]

I have this piece I need to attach at the top side of the cpu for the NH-D15 cooler, I attached this at the -7mm setting, this was advised to me for AM5, but it seems I do not need to use the offset and just attach it at 0 ?

image link

 

[hardware_bxl]

Also the thermal grizzly comes with a spreader, do you need to use that maybe for this particular thermal paste?
I appreciate your experts advice, I know thermal paste is probably not that critical, but I like to do it the right way 🙂

image link

 

[COLGeek]

When spreading, you want as little compound as possible (on the CPU itself). Spread evenly and quite thin.

Use the offset, it should help. I have the same cooler in an AM5 and ran into the same issue.

 

[hardware_bxl]

@COLGeek: yes I just read about the offset being better for AM5.
Would you advise me to use the Noctua own thermal paste, or the one I bought, the Thermal Grizzly Kryonaut?
And do you generally prefer spreading it or is a drop in the middle just as efficient?

 

[COLGeek]

I have used the thin film method for decades. Does the job and minimizes waste/squeeze-out. Less is more when you consider you are filling very small spaces when you properly tighten the heatsink itself to the cooler.

Either thermal compound is fine. I use the Noctua stuff often. Just make sure ALL of the previous stuff is removed before applying something else, especially if changing to a different product.

 

[hardware_bxl]

Well, this is far from the thin film you were talking about, but i was so afraid i would not cover it all. Anyway, I left the 7mm offset as suggested and re-installed the cooler.
I had not mentioned yet that I replaced the 7900x with an 9800x3d cpu, but that I did (as you can probably not see below the layer of thermal paste), and after cinebench 2024 running on multicore, i still was surprised by the high temperatures, but I think they are ok-ish as I read others also showing this.
Here are some stats:

(max values)
CPU (Tctl/Tdie) - 96.1C
CPU Die (Average) - 95.0C
CPU CCD1 (Tdie) 98.8C
-> Core Temperatures 94.7C
-> L3 Temperatures 66.0C
CPU IOD Hotspot - 62.2C
CPU IOD Average - 56.4C
CPU VDDCR_VDD VRM (SVI3 TFN) - 50.1C
CPU VDDCR_SOC VRM (SVI3 TFN) - 46.0C
CPU VDD_MISC VRM (SVI3 TFN) - 47.0C
Core VIDs 1.265V
Core Clocks 5,225.0 Mhz
CPU Package Power 137.567W

Cinebench score on multicore was 1274, i think that is expected, little on the lower side.

Leaving it idle for 5 minutes after and temps drop fast to:
46.4, 46.1, 33.8, 30.0, 31.0, 46.5, 41.7, 37.3, 34.5 and 37.0C

This is all normal and would suggest I did a decent enough installation job? So unsure lol.

image link

 

[Roland Of Gilead]

Well, this

It's not perfect, and still IMO, a little too much. But it's better than the first pics. The first pics you showed look like the past has pumped out a bit. That could just be down to the application though.

 

[hardware_bxl]

Yes, I know, I was just afraid putting too little and not cover the whole area and that spreading was not going that smooth as I hoped 😎.

Any comments from any experts on the temps, is this to be expected from a 9800x3d with the Noctua NH-D15 Chromax cooler?

 

[SyCoREAPER]

It wouldn't have prevented the paste issue but personally I would get a contact frame. Ensure there is even pressure and contact as well as a level mating surface for the cooler.

 

[Roland Of Gilead]

The 9xxx series do run warmer. IMO, your idle temps are about where they should be. Yes, in recent gen AMD lets say the 5xxx series, you could expect lower idle temps. Mine hovers around 30-35c. But thats because currently it's summer here. Ambient plays a part, so you can expect temps to be about 10-15 over ambient. Currently it's 17c here (Yeah, that's an Irish summer for you!_

 

[hardware_bxl]

@SyCoREAPER : I did get the contact and sealing for AM5 from Thermal Grizzly (you can see it around the CPU in the last picture I shared).

@Roland Of Gilead : Ok, that somehow comforts me, but I was still surprised by the high 98.8C peak on the CPU CCD1 (Tdie) but I guess it's drawing 137W too, but if you think about throttling, that's where this would happen right around those temperatures? I guess it can differ from chip to chip and some will run hotter than others too? Of course it's a multi-core test on Cinebench, so even when full gaming this stress level won't often be reached I guess? Temps when idle stay decent around 44C for the CPU die, ambient here is about 22C as it's summer in Brussels too although we have rain today.

 

[dwd999]

You can't discuss and compare temps without explaining the various factors such as fan curves/speeds, case fan setup, curves and speeds, and even ambient temps. Were the cpu fans running at 100 percent, the case fans at 100 percent, etc. when you experienced those temps.

 

[hardware_bxl]

I did read about underclocking to avoid those high-high temps, but I do think in real life scenarios, the multi-core stress will be sporadic and if this happens, the fans will just spin up and that's about it. I guess I could create a fan curve to only really start at 70C and this would be sufficient?
But overall I could be happy (enough) with my installation, the offset of 7mm i still better to leave this, the thermal paste is decent enough, the noctua cooler does its job?

 

[hardware_bxl]

@dwd999 : ah, the cpu fan went up to 1,405RPM, the case fans not really increased and remain at 950RPM each..(the other temps were not higher, so that makes sense the case fans not sped up) CPU fan in idle is at 575RPM

 

[dwd999]

You could try manually setting every fan that you have at 100 percent and see what temps that gives you.

I use FanControl to test for things like that:

Fan Control - A highly focused fan controlling software for Windows

Fan Control is a free software that allows the user to control their CPU, GPU and case fans using temperatures.

getfancontrol.com

 

[SyCoREAPER]

@SyCoREAPER : I did get the contact and sealing for AM5 from Thermal Grizzly (you can see it around the CPU in the last picture I shared).

@Roland Of Gilead : Ok, that somehow comforts me, but I was still surprised by the high 98.8C peak on the CPU CCD1 (Tdie) but I guess it's drawing 137W too, but if you think about throttling, that's where this would happen right around those temperatures? I guess it can differ from chip to chip and some will run hotter than others too? Of course it's a multi-core test on Cinebench, so even when full gaming this stress level won't often be reached I guess? Temps when idle stay decent around 44C for the CPU die, ambient here is about 22C as it's summer in Brussels too although we have rain today.

No you didn't. Go look it up.

 

[ohio_buckeye]

Personally I’m not a fan of thermal grizzly. I like Arctic MX-4. Do a drop in the center, spread with a finger. Been doing it that way for years.

 

[COLGeek]

Just for reference, you likely put 2x what you needed on the CPU. You'll just have a bunch of squeeze out around the IHS nooks and crannies.

 

[hardware_bxl]

@SyCoREAPER : here you can see the thermal grizzly contact & seal for am5.

@ohio_buckeye : it seems the thermal grizzly kryonaut does not spread easily and since I am not that experienced and afraid of doing it wrong, I wanted to see the paste covered on the cpu and then automatically you do too much... maybe must re-apply it one more time and use the noctua thermal paste or buy artic mx-4 like you say, mx-6 is worse btw?

@COLGeek : yes I know... just not sure and the grizzly spreads difficult in my opinion...

 

[SyCoREAPER]

@SyCoREAPER : here you can see the thermal grizzly contact & seal for am5.

@ohio_buckeye : it seems the thermal grizzly kryonaut does not spread easily and since I am not that experienced and afraid of doing it wrong, I wanted to see the paste covered on the cpu and then automatically you do too much... maybe must re-apply it one more time and use the noctua thermal paste or buy artic mx-4 like you say, mx-6 is worse btw?

@COLGeek : yes I know... just not sure and the grizzly spreads difficult in my opinion...

That it is, however what it isnt is in your first post of pictures. Those only show the standard MB retention mechanism.

And thermal grizzly paste is benchmark paste, not daily. It dries up faster and needs to be reapplied more frequently unlike something like Noctua NH2

 

[ohio_buckeye]

@SyCoREAPER : here you can see the thermal grizzly contact & seal for am5.

@ohio_buckeye : it seems the thermal grizzly kryonaut does not spread easily and since I am not that experienced and afraid of doing it wrong, I wanted to see the paste covered on the cpu and then automatically you do too much... maybe must re-apply it one more time and use the noctua thermal paste or buy artic mx-4 like you say, mx-6 is worse btw?

@COLGeek : yes I know... just not sure and the grizzly spreads difficult in my opinion...

I'm just saying that's what I've been using the last couple of years or so. I think really when you look at thermal paste, many of them are pretty close. If you already have Noctua paste, roll with that. I just keep a little MX-4 around.

 

[MEMOFLEX]

The 9xxx series do run warmer. IMO, your idle temps are about where they should be. Yes, in recent gen AMD lets say the 5xxx series, you could expect lower idle temps. Mine hovers around 30-35c. But thats because currently it's summer here. Ambient plays a part, so you can expect temps to be about 10-15 over ambient. Currently it's 17c here (Yeah, that's an Irish summer for you!_

My 9800x3d idles in the high 30's with a -15 all core negative offset on a 360mm AIO. Ambient not that much different here in Bristol, UK.

An hour long cinebench load wont push the temps past 70 degrees though so mine operates in quite a small temperature delta no matter what the workload.

O11D Evo case with all Corsair link fans set to 50% set speed

Edit : I use mx4 and the thin spread approach

 

[hardware_bxl]

An hour long cinebench load wont push the temps past 70 degrees though so mine operates in quite a small temperature delta no matter what the workload.

That makes me just sceptic, not because of what you are saying, but that would basically imply that either I did a very bad job with my installation, or the temperatures differ so much from chip to chip, that would almost make me want to return my 9800x3d, but I am sure they will deny the RMA...
I know you have an AIO installed, but the Noctua cooler is claiming to be one of the best air coolers, so that would still not make me happy seeing these 98C temps at full multi-core load... :-(