A Shelf-type Noctua HSF (strong XP-90 lookalike)

[Felger Carbon]

Here.

If you follow the link to Noctua, and then follow the link to the english-language review (dated today!), you will find some cooling performance figures with no fan that I find very hard to believe...

Direct link to cooling data here (no-fan data towards bottom of page).

[Modo]

Well, I'd believe it. It looks like an improved version of the TR SI-128 SE. More fin area, more space between fins, more heatpipes, so less need for added airflow. It should perform very well with and without a fan.

And if you've followed, the big tower heatsink from Noctua also shows very good performance. So I'd assume they just keep consistent with the initial high-quality design.

[Bluefront]

Boy....that looks very similar to an XP-120, just taller. And I like their explanation of the relatively wide fin spacing....."a relatively wide fin spacing that guarantees optimal heat dissipation even at low fan speeds".

Far as I'm concerned....the wide fin spacing was necessary because of the poor performance of the 12P fan when confronted with restriction. None-the-less, looks pretty good.

[walle]

Felger, thanks for the heads up.

This looks promising, perhaps finally a blow-down heatsink really worth its name?. I like the wide fin spacing, but I particularly like the gaps on each side of the heatsink even more (it suggests that this unit would actually be somewhat able to cool the board) something, which the Scythe Andy Samurai Master was so utterly inferior at, but to be fair; the design didn’t suggest such qualities to begin with.

Regarding the Thermalright SI-120 SE, well, I’ve used that one too and it performed better than the Andy Samurai in my experience, yet, it didn’t cool the motherboard any better, but to be fair, they never claimed it would, as far as I can remember.

I actually like this latest creation from Noctua, lets hope that it performs.

Bluefront,

We know that you’re still royally pissed having spent premium money on the NF-P12 fan getting disappointed that it didn’t perform as you expected, fair enough, but I also own a couple of these fans and even thou I agree with you that they do not perform exceptionally well when confronted with restriction, they do not perform as bad as you would like to have it, so no need go maligning them shunting them as the plague now.

By the way, did you actually travel to Austria giving Noctuas engineers a clout around their ears? if you didn't, but plan to, I’ll bloody well join you smacking them around some, besides, the alps are nice this time a year


Cheers

[Bluefront]

walle...... it's not the cost of the fan that pisses me off so much as the continual praise of this fan, lauding it's false ability to excel in restrictive situations. Hell there are cheap Scythe fans that do as well....probably better, and cost 1/2 as much.

[FartingBob]

Load temps i guess maybe is possible if its a cold day and on a open testbed. That idle temp though (22.5c!) shows that the review was clearly done in a very cold room, or the temps were measured after a cold boot in BIOS when the thing has been given about 20 seconds to warm up..

[walle]

walle...... it's not the cost of the fan that pisses me off so much as the continual praise of this fan, lauding it's false ability to excel in restrictive situations. Hell there are cheap Scythe fans that do as well....probably better, and cost 1/2 as much.

Funny, your statement mirrored my view exactly whereas I though my previous statement reflected yours (as perceived by previous inputs), anyway, the parroting praise these fans have received is somewhat disturbing and don’t speak highly of those review sites having praised them either. Perhaps I could have overlooked that to some extent hadn’t the price of these fans been premium.

[walle]

Load temps i guess maybe is possible if its a cold day and on a open testbed. That idle temp though (22.5c!) shows that the review was clearly done in a very cold room, or the temps were measured after a cold boot in BIOS when the thing has been given about 20 seconds to warm up..

I don’t think anyone reading the review would take those temp readings as that accurate nor to serious Bob, I might be mistaken of course assuming that just because we don’t, others wouldn’t.

Disclaimer:

With reservation for those whom are new to computers lacking experience.

[elpibe10]

Nice looking CPU heatsink.

I'm definitely going to get one if it performs better than TR SI-128SE.

... hope to see a review here soon


Welcome to my WiNDy House of Soldam :

http://www.freewebs.com/elpibe10/

[bgavin]

I saw that new Noctua review today. The tester said 19C was the ambient temperature.

I've never been able to get any of my processors to idle 1.5C higher than ambient. Even with a lot of air flow through them, let alone no fan at all.

The new Noctua has half the heat pipe capacity of the Zipang, so I really do have to question those numbers. After reading a number of his reviews, he is indeed a big Noctua fan. Oops, bad pun.

I like most of the new Noctua construction. Assymetric heat pipe side (one side only) is not something I like, but the base does contribute to holding up the fin assembly. I won't have a sink that supports the fins entirely by heat pipes on one side only.

What I would like to see is some common fan testing. SI-128 SE, XP-120, Zipang, the new Noctua, etc, etc, all with the same fan, same processor, and same board.

[Modo]

Load temps i guess maybe is possible if its a cold day and on a open testbed. That idle temp though (22.5c!) shows that the review was clearly done in a very cold room, or the temps were measured after a cold boot in BIOS when the thing has been given about 20 seconds to warm up..

Or, the used monitoring software has wrong reference values (old SpeedFan or CoreTemp, for example).

[Felger Carbon]

What I would like to see is some common fan testing. SI-128 SE, XP-120, Zipang, the new Noctua, etc, etc, all with the same fan, same processor, and same board.

Commonality in testing is such a great idea that nobody would argue against it, right?

Uh, which can get from Long Beach Harbor to Avalon on Catalina Island faster, an outboard-powered rowboat or an Air Force F-15 fighter jet? For commonality, both have to travel at the same altitude. Outboard chug chug, F-15 glub glub (altitude minus one foot sea level in each case).

I like both the GW NCB and Slipstream fans. The SI-128SE will work better with the former, a Ninja with the latter. Which of the two fans do we use for commonality testing?

[Cistron]

The new Noctua has half the heat pipe capacity of the Zipang, so I really do have to question those numbers. After reading a number of his reviews, he is indeed a big Noctua fan. Oops, bad pun.

Quality > quantity, I'd say. After all, a lot will depend on how well the heatpipes are soldered into the base and how good the contact between heatsink and heatspreader is.

[NeilBlanchard]

Hello,

When I first saw this heatsink, I wondered why there is a section of fins "missing" on the sides. My hypothesis is this is the section where the fins are nearly parallel with the trailing edges of the fan blades -- and this might be causing more noise, and doesn't gain enough cooling to be worth it. It also might allow more air to reach the motherboard; though the utility of this is probably minimal.





I like the wide(r) fin spacing, and I like the "wavy" leading edges -- I think this makes it quieter.

[bgavin]

Quality > quantity, I'd say. After all, a lot will depend on how well the heatpipes are soldered into the base and how good the contact between heatsink and heatspreader is.

I would be very surprised if the various manufacturers did not buy heat piping from the same manufacturer. 6mm pipe is 6mm pipe. Both solder the heat pipes to the fins, and solder to the base.

The Noctua has six segments connecting base to fins, where the Zipang has 12 segments. If the Noctua cooler does indeed outperform the Zipang, it does it with less fin area and half the heat pipe segments.

As for common fan testing, pick one. I'm easy. It would be interesting to see how Fan X performs on different coolers.

[Felger Carbon]

When I first saw this heatsink, I wondered why there is a section of fins "missing" on the sides.

The stated reason is to improve the cooling of the mobo, as you mentioned. Another, possibly more important, reason is to allow a better/more compatible mounting to be used, allowing room for a screwdriver.

The new 3RSystems Prima/Sunbeam/SilenX has a wasp-waist instead of the rectangular "missing" section. This is why the new heatsink can mount in the correct direction on most AMD boards. Again, the wasp-waist allows a screwdriver to reach the mounting hardware.

As an AMD fan, I applaud!

[NeilBlanchard]

Hello,

At 550gm (without fan), it is lighter than the Xigmatek -- nice.

[Redzo]

walle...... it's not the cost of the fan that pisses me off so much as the continual praise of this fan, lauding it's false ability to excel in restrictive situations. Hell there are cheap Scythe fans that do as well....probably better, and cost 1/2 as much.

Well said !
I have one of them too and I am NOT impressed with this fan. any one of mine black and white Nexus fans gives me better temps and consistent airflow for a much cheaper price.

[nyu3]

From this picture, http://www.overclock3d.net/gfx/articles ... 13369l.jpg , it looks like the practical reason for the missing fin sections is to let the screwdriver fit through.

Noctua heatsinks are overpriced here in Canada Ninja and Ultima-90 can be had for <$38 , HR-01 plus is only $44. Noctua NH12P is $55

I'd be interested in a SilentPCReview update of the HR-01 (the Plus version adds 2 heatpipes, fan clips, and bolt mounting for LGA775). SilentHardware.de shows that the HR-01 Plus outperforms Xigmatek HDT1283, Ultra-120 Extreme, and Ninja Plus with the 800rpm S-Flex fan.

[Modo]

Noctua heatsinks are overpriced here in Canada

Not really. Those are their normal prices. Same with fans. Good quality at high prices. Sort of like paying for a logo with a fruit on some computers.

[colm]

believing in reverse cooling, no fan on the heatsink, ductwork. Acheiving static pressure is a challenge. the xp90 meets that challenge. I have to fail the noctua

was quite intrigued at first glance...then found this disappointing photo.


bad opening

and furthermore...as even the spcr article about fans has learned about 92mm fans being more common on cpus, more likely to have a very large rpm range, like the cpu fan pwms on alot of mobos utilize...
I have to fail again, for the lack of precisely 92mm doing its cpu pwm job.

once again, a sexy heatsink with a hidden backdoor...
a 120mm hardly any fan mamager, and other unique scenarios would use this quite ok. I like to use whatever I got from oem to the max before really hacking into cpu fans and sinks. today,it is unlikey to find such a problem, making that darn hole even bigger....

[zenzero-2001]

Had some problems understanding your post.

I thought the noctua uses a 120mm fan not a 92mm ?

I think the gaps on the heatsink are to aid cooling of the motherboard components.

believing in reverse cooling, no fan on the heatsink, ductwork. Acheiving static pressure is a challenge. the xp90 meets that challenge. I have to fail the noctua

was quite intrigued at first glance...then found this disappointing photo.

http://i227.photobucket.com/albums/dd25 ... c12p_2.jpg
bad opening

and furthermore...as even the spcr article about fans has learned about 92mm fans being more common on cpus, more likely to have a very large rpm range, like the cpu fan pwms on alot of mobos utilize...
I have to fail again, for the lack of precisely 92mm doing its cpu pwm job.

once again, a sexy heatsink with a hidden backdoor...
a 120mm hardly any fan mamager, and other unique scenarios would use this quite ok. I like to use whatever I got from oem to the max before really hacking into cpu fans and sinks. today,it is unlikey to find such a problem, making that darn hole even bigger....

[Erssa]

Had some problems understanding your post.

I thought the noctua uses a 120mm fan not a 92mm ?

Yes, it uses 120mm NH-P12 fan.

Noctua heatsinks are overpriced here in Canada

Not really. Those are their normal prices. Same with fans. Good quality at high prices. Sort of like paying for a logo with a fruit on some computers.

walle...... it's not the cost of the fan that pisses me off so much as the continual praise of this fan, lauding it's false ability to excel in restrictive situations. Hell there are cheap Scythe fans that do as well....probably better, and cost 1/2 as much.

Funny, your statement mirrored my view exactly whereas I though my previous statement reflected yours (as perceived by previous inputs), anyway, the parroting praise these fans have received is somewhat disturbing and don’t speak highly of those review sites having praised them either. Perhaps I could have overlooked that to some extent hadn’t the price of these fans been premium.

Noctua are a bit expensive, but I don't think all the hate Noctua gets is justified. The fan itself is no more expensive then Nexus was 2 years ago. For example fan prices in Finland:

Noctua NF-P12 19.90€
Scythe Slip Streams 15.90€
Scythe S-FLEX 18.90€
Nexus Real Silent Fan 16.90€

You can find the Nexus a bit cheaper then 16.90€, if you look from the right places. But the rest of them are just about as expensive as Noctua, no matter where you look. I paid 12e for my second Nexus last year, I also bought silicon fan mounts, that cost 3.90€ and Fan Mate 2 for 4€, because my motherboard can control only the cpu fan. Noctua isn't so expensive, when you consider it comes with silicon fan mounts and their fan adapter. In total I paid the same price for my Nexus. Granted, now I have a better fan controller and I'm set for life with those silicon fan mounts...

As for P12s ability to fight against restrictive situation. It should perform well enough for just about every user except bluefront and his crazy fan filters. Rpm for rpm, it beats Slip Stream in airflow based on every heatsink test I have seen. Although I haven't seen any tests below 800 rpm, so it's could be a poor performer at low rpms, but every test indicates it has no problem with restrictions with 800+ rpms. But anyway I'm going save my final judgment until SPCR test them. Those dragonmods results look bogus. Especially those passive temps. It takes away whole credibility from the test imo. Hopefully Noctua sends SPCR this heatsink.

Even though it's crazy expensive atm (our biggest e-tailer charges ridiculous 83€ for it!), review would be interesting because it's 1mm shorter (with fan) then Mini Ninja, which means it might be the heatsink for HTPCs.

[bgavin]

FrostyTech just did a review on the new Noctua. Using either the Blue or Black resistor block, the Zipang is both quieter and cooler than the Noctua. At full fan speed, the Noctua cools better, but with much more noise.

Frosty noted the base on the Noctua is convex in one plane, and concave in another. I find this disturbing for a premium sink. Yes, it can be lapped, but at this price point, should it have to be? (Yes, the TRUE is more $$$, and it is concave also.)

[Erssa]

Those dragonmods results look bogus. Especially those passive temps. It takes away whole credibility from the test imo.

I actually bothered to do a bit of digging and found out, that DragonSteelMods is the official US launch partner for Noctua. No wonder those numbers looked so weird...

[Modo]

@Errsa
I have nothing against Noctua adding a premium for their products. I used the Apple reference to point out how futile rants against high prices of high-end products are. Cheers.

Frosty noted the base on the Noctua is convex in one plane, and concave in another. I find this disturbing for a premium sink. Yes, it can be lapped, but at this price point, should it have to be?

Noctua claims the heatsink base doesn't need to be flat because of the thermal compound they use. FrostyTech says they use a standard thermal compound on every heatsink. Whether that compound is inefficient on this surface, or whether Noctua's claim is wrong, I don't know. But I do wonder why FrostyTech would deliberately not use the thermal compound recommended and included by the manufacturer.

[NeilBlanchard]

Hello,

The best situation is for close physical contact, which means both surfaces should be flat, ideally. Flat and relatively smooth -- these are different things. The thermal paste is there to "make up for" imperfect smoothness -- but it can't really "make up for" a non-flat surface. That, and the differences in various thermal pastes are generally less than many folks think...

[Modo]

I see that, but even a 2 degree load difference can throw off the results quite a bit, especially if you apply that uncertainty to an unknown number of the presented results. This is about heatsink testing in general. YMMV and all that.

[nick705]

Those dragonmods results look bogus. Especially those passive temps. It takes away whole credibility from the test imo.

I actually bothered to do a bit of digging and found out, that DragonSteelMods is the official US launch partner for Noctua. No wonder those numbers looked so weird...

ugh... I remember a while ago reading a DragonSteelMods review of the original NH-U9F heatsink, and thought at the time it was suspiciously gushing.

This sort of thing does no-one any favours - certainly not the readers, and apart from losing the "review" site any credibility whatsoever, it reflects badly on Noctua themselves, whose products seem otherwise generally good enough to stand on their own merits.

It makes life so much harder when you have to always read between the lines for ulterior motives, but I suppose that's a rant for another day.

[velvet45]

Modo,
I would also add here that it would only make sense for a review site (Frosty tech) to keep the same thermal compounds for all sinks reviewed to reduce sample variance.
If Noctua or some other manufacturer wants to claim improved performance, they should not rely on "special" considerations such as using a certain compound to make up for imperfections in manufacturing. Unless, that is... they are competitive enough in their pricing to provide that extra performance for no extra added cost.
That stuff about a special surface (convex) or less than perfectly flat surface sounds rather odd.
The proof is in the performance ...period- all other claims made by manufacturers are meaningless.