Noctua changed coldplate quality, or just a one-off?

[tabascosauz]

So I went and took the new L12S off, and ordered a L9x65SE because the fitment wasn't going to work out. In the process of cleaning it up, I noticed that the L12S' coldplate was......odd.

I've always been used to Noctuas (minus L9i and L9a, different coldplate) having the same polished finish that falls just short of a mirror finish. As in, there's always the telltale, uniform, concentric ring-like machining marks but otherwise it's smooth enough to see a reflection.

This L12S is completely different. It isn't smooth at all, with a consistently almost "matte" finish. Frustratingly, it really tears up the 99%IPA-soaked paper towels I always use to clean my coolers - leaves a lot of dust/bits/lint behind with every wipe. Exactly what you'd expect if you rubbed a wet paper towel over a polishing stone, for example.

--

L12S: No reflections at all, partly due to the roughness, partly due to the way it's machined around the centre point:


D9L: ~6 years old. Kinda distorted, has parallel "waves" running through it in a west-east direction. But still has a reflection and the machining rings:


U9S: ~1.5 years old, the smoothest coldplate I have from Noctua. Near-mirror, smooth, and flat unlike the D9L, but still has the telltale rings:


--

As should be relatively obvious from clicking on the full resolution photos, the L12S is not like the others, in a rather significant way. Did I miss some major press release stating that Noctua changed its tooling for making these?

It honestly reminds me a little bit of textured plastic, just on a much smaller scale of course:


Anyways, I don't care for this one because I'm returning it for a L9x65. But it was strange to see the finish on a Noctua product depart noticeably from any other Noctua I've owned.

Anyone with a recent L9x65/D9L/U9S/L12S/L12/U12S/U12A/U14S/C14S/D15/D15S feel free to chime in if you remember what your coldplate looks like.

 

[the54thvoid]

Maybe contact Noctua direct? Get the response from the horses mouth? What little I know about heat transfer; the smoother the surface, the more complete the contact and conduction of heat. That surface looks mottled - wouldn't that create pockets of non-uniform cover? Looks almost... fake Noctua.

 

[tabascosauz]

Maybe contact Noctua direct? Get the response from the horses mouth? What little I know about heat transfer; the smoother the surface, the more complete the contact and conduction of heat. That surface looks mottled - wouldn't that create pockets of non-uniform cover? Looks almost... fake Noctua.

Yeah I think I'll drop em a line. Mottled is definitely the word I'm looking for. It's the kind of finish I'd expect out of a 212 Black (did one for a friend not long ago). As for thermals, it did do better than my push-pull U9S which was a little surprising, but it was on the testbench and I wasn't trying very hard for a scientific comparison. So if there are any thermal penalties, I don't think I'll be able to see the difference with the MX-4 that I'm using.

Makes me wonder if it's a one-off that somehow slipped through the final polishing stage without QC catching it. Doesn't look like there's anything wrong with the coldplate itself, just that someone forgot to put it under the polishing machine.

--

Sent an email off to Noctua. In the meantime, I managed to get better photos:



L6x65SE should be here on Thursday, I guess we'll find out if this was just a one-off.

 

[ThrashZone]

Hi,
Yeah looks like their trying to catch up to water blocks to fit better with concave cpu's
Not a bad if amd or intel were consistent but intel sure isn't chip series to chip series lol

7900x & 9940x are like bowls
10900k is pretty darn flat so big differences for a air cooler to match seeing there isn't flex on them is there ? water blocks do flex a little.

 

[MIRTAZAPINE]

The D15S I used to have is got a mirror like finish to it. I can't comment for the other noctua though. I don't think this would affect much, flatness is the number one thing before smoothness. though in general Noctua base is rather slightly dome shape convex.

 

[Jetster]

Just bought a NH U12S and it looked good

 

[cst1992]

Just bought a NH U12S and it looked good

How's the mounting mechanism? Is it good like it's shown to be?
I wanted to get one of those, but it was too expensive (INR 10,000/$138) where I am, so I got a Gammaxx 400 with the stupid plastic push-pins.

 

[P4-630]

I have lapped some coolers in the past, since I'm using noctua I haven't done that anymore.

 

[qubit]

That's either a faulty plate, Noctua have gone low quality (unlikely) or a fake. Good idea to talk to them about it.

 

[Jetster]

How's the mounting mechanism? Is it good like it's shown to be?
I wanted to get one of those, but it was too expensive (INR 10,000/$138) where I am, so I got a Gammaxx 400 with the stupid plastic push-pins.

It's all quality. Easy install, way easier than the CM 212. I also did a test against the 212. It out performs if by about 10% Not sure if that makes it worth twice as much but its a good looking cooler with solid mounting

 

[cst1992]

a faulty plate

Most likely. I've yet to see a fake Noctua anyway.

It's all quality. Easy install, way easier than the CM 212. I also did a test against the 212. It out performs if by about 10% Not sure if that makes it worth twice as much but its a good looking cooler with solid mounting

Okay, might consider it for my next build.

 

[Chrispy_]

to fit better with concave cpu's

Likely explanation.

Circular machining from the centre of each plate is definitely more expensive in terms of tooling setup and machining time. Noctua probably wouldn't move to a more expensive, less-efficient process without some credible reason.

The rougher finish may or may not be intentional. The need to lap to a perfectly-flat, mirror finish vanished with the rise of the IHS. Also, while physics dictates that the thinnest possible layer of paste provides the best heat transfer, the thinnest layers of paste also dry up to the point of uselessness within 12 months. Put a bit of roughness on there and it'll stay marginally thicker and still be liquid when removed for cleaning four years later. The extra 0.2C temperature increase is worth it for the reduced maintenance.

 

[DeathtoGnomes]

subbed, waiting to hear what noctua has to say.

 

[ThrashZone]

Likely explanation.

Circular machining from the centre of each plate is definitely more expensive in terms of tooling setup and machining time. Noctua probably wouldn't move to a more expensive, less-efficient process without some credible reason.

The rougher finish may or may not be intentional. The need to lap to a perfectly-flat, mirror finish vanished with the rise of the IHS. Also, while physics dictates that the thinnest possible layer of paste provides the best heat transfer, the thinnest layers of paste also dry up to the point of uselessness within 12 months. Put a bit of roughness on there and it'll stay marginally thicker and still be liquid when removed for cleaning four years later. The extra 0.2C temperature increase is worth it for the reduced maintenance.

Hi,
EK magnitude is also pretty rough circle milled cold plate, looks like noctua was watching and machined the same way ek did but looks like noctua did it better from the image on this op compared to the tpu review plus I have one copper ek cold plate and it is or was pretty darn ribbed and was a pita to get all thermal paste off lol until I buff it smooth

EK-Quantum Magnitude CPU Water Block Review

EK Water Blocks finally launches the EK-Quantum Magnitude, a halo CPU block with customization cues from the EK-Supremacy EVO from yesteryear that kicks things up a notch. With a clear focus on design, material choice, performance, and customization, it aims to offer a lasting CPU cooling...

www.techpowerup.com

subbed, waiting to hear what noctua has to say.

Already did I misplaced where exactly. here it is

(Anti) SFF fun house

@Mussels never said their the best or non-plus-ultra. so: show me 1 or 2 more cases from "name" brands (on the market for at least 15y) that have a midi tower with TG (not just a cut out in panel), space for 140 sized fans all around (rear/top/front/bottom, and space for two 120 fans on the...

www.techpowerup.com

 

[R00kie]

Hi,
EK magnitude is also pretty rough circle milled cold plate, looks like noctua was watching and machined the same way ek did but looks like noctua did it better from the image on this op compared to the tpu review plus I have one copper ek cold plate and it is or was pretty darn ribbed and was a pita to get all thermal paste off lol until I buff it smooth

EK-Quantum Magnitude CPU Water Block Review

EK Water Blocks finally launches the EK-Quantum Magnitude, a halo CPU block with customization cues from the EK-Supremacy EVO from yesteryear that kicks things up a notch. With a clear focus on design, material choice, performance, and customization, it aims to offer a lasting CPU cooling...

www.techpowerup.com

Already did I misplaced where exactly.

Thats interesting, my 5 year old Supremacy MX is mirror finish, have they started cheaping out on their new blocks then?

 

[ThrashZone]

Thats interesting, my 5 year old Supremacy MX is mirror finish, have they started cheaping out on their new blocks then?

Hi,
Not sure it's cheaping out just trying new things
If ek were cheap they'd just keep doing the same thing and charge more for it
Now at least it's new design and charge a lot more lol it's stunning how much more the magnitude line costs.

 

[LTUGamer]

I believe that it is even impossible to measure temperature difference between these designs. Since it is impossible to put thermal coumpound 100% identically, that make bigger difference than those coldplate designs.

 

[ThrashZone]

I believe that it is even impossible to measure temperature difference between these designs. Since it is impossible to put thermal coumpound 100% identically, that make bigger difference than those coldplate designs.

Hi,
AMD might be difficult seeing they that I've noticed don't shore each cores temperature
Intel it's very easy to see a difference if a cooler mates well with a chip because the core temp spread which is hottest to coolest core temperature each core has it's own sensor
So on a 10 core or 18 core processor if the core temp spread is within 8-10c from hottest to coolest core you're mating pretty darn well.
Most on old coolers are well over 10c apart showing how poorly it's mating.

Here's some tests I did a while back showing just how well these two water blocks handles core temp spread optimus clearly did a better job of maintaining a tight core temp spread.

 

[tabascosauz]

Noctua confirmed that apparently the L12S is supposed to look like this:



Having a bit of a back and forth with them rn trying to figure out which coolers are like this.



Nice touch that even the email is in brown first time in 8 years I seen it

 

[qubit]

Sounds like lower quality manufacture, one of my possible options, above and the one I thought least likely of Noctua.

 

[tabascosauz]

Sounds like lower quality manufacture, one of my possible options, above and the one I thought least likely of Noctua.

To be fair, I don't care about the cooling and difference that an unpolished base makes. The L12S is not really a performance cooler by any means but still cools just fine, and from what I can tell the big boys like my C14S are still true to the Noctua name.

I just hate how hard it is to clean with paper towels. Might have to start using an old microfiber cloth to give it a final wipe after every cleaning.

 

[80-watt Hamster]

To be fair, I don't care about the cooling and difference that an unpolished base makes. The L12S is not really a performance cooler by any means but still cools just fine, and from what I can tell the big boys like my C14S are still true to the Noctua name.

I just hate how hard it is to clean with paper towels. Might have to start using an old microfiber cloth to give it a final wipe after every cleaning.

My Cryorig C9 is the same way. Regardless of cooler type/brand, I do my final cleaning passes with coffee filters: w/alcohol first, then dry. Wet filters can leave some lint, but dry ones are for all practical purposes lint-free.

 

[ThrashZone]

Hi,
Yeah ek magnitude cold plate was tough to get the paste off
Gets into all the low spots it really was terrible.

 

[Totally]

Hi,
Yeah ek magnitude cold plate was tough to get the paste off
Gets into all the low spots it really was terrible.

Just drop it in hot soapy water, it comes right off.

 

[ThrashZone]

Just drop it in hot soapy water, it comes right off.

Hi,
Don't think I have that much slack in the hosing
Buffed issue away image about 11 posts up