FRONTPAGE Noctua NF-A12x15 Fan Review

Noctua has released a new 15 mm thick fan, named the NF-A12x15. It is designed to fit in skinny places and to provide relatively high static pressure. They faced challenges getting a fan this thin. For example, “For the 12cm NF-A12x15, a steel-reinforced motor hub and brass axle mount are being used in order to assure the required stability despite the fan’s 15mm "thin profile.” Remember, Noctua guarantees their products out to six years and to make a thin fan last that long took some work. They also never made a fan this thin. As our press contact commented, “All of the new models are new form factors for us.” So let’s have a look at this, the newest entry into the slim fan derby:

Click to expand...

Click here to view the article.

Sadly i cannot understand your graphs. did you guys happen to do any real world testing. i saw a video of their booth at computex and it showed that 1 of these fans beats out 2 nf-f12 fans on an air cooler.

they had a graph that showed this fan beats the s12 in a case (but s12 wins with no restriction). and it beats the f12 when used on a cooler or radiator (but the f12 wins with 100% restriction). basically saying in real world usage this is the best fan period. i trust and love noctua (just not their color). so i have to believe them.

i would like to see some real world test, maybe a 3 scatter plot graphs that have airflow on x axis, and noise on y axis. first graph could be the fans on a radiator (push) and the 2nd graph could be case intake through a mesh grill. 3rd, could be temp on x axis and noise on y axis for radiator (push).

this would show how well it competes with the s12 and f12 along with other manufactures best fans (as far as noise and airlflow/cfm).

i ordered 6 to replace all my fans, so i guess im in it either way.

thanks for the info!

in case anyone wants to see the video, search youtube for "Noctua - Even Better Than Before | Computex 2017" a video by CRIT.

These were Real World Tests (tm). I did cooling tests on a heatsink with fans back in 2009-2010. Those tests did not differentiate fans enough. Sometimes there was only a fraction of a degree between fans, and the SPL testing was open to many vagaries.

Now that I have anemometers I can test airflow directly. These are not tests where you see fans matched by RPM. That always seemed artificial to me. These charts are set up so that the SPL, airflow under the test condition and airflow unobstructed (a non-realworld testing point, useful only in comparing to specification), are set up so you can compare various fans to each other under three realworld conditions: through a dust filter (for case fans), through a 16 FPI rad (a proxy for the average heatsink) and through a 30 FPI rad.

You basically pick your fan usage, find your fan and compare it to others.

The gravamen of the charts is the middle column, the one that changes color as you step through the charts. You just have to spend some time looking at them, letting the data speak to you. You cannot just look at the charts and pass on. There is too much information for that.

In the second chart you can see the NF-A12x15 and the NF-F12 side by side. The two fans were dead even in airflow with the through the 16 FPI rad. The NF-A12x15 was 3 dB quieter than the NF-F12 (in unobstructed airflow the NF-F12 was better; shows how non-realworld that measurement is). When pushing through the 30 FPI rad, the third chart shows you that the NF-A12x15 was 6 CFM better than the NF-F12.

In my reviews the test conditions are stable. I use the same measuring tools, the same filters and the same rads for the fans. You haven't said which heatsink Noctua was using to show off this excellent fan. I suspect that it was a tight heatsink, presenting more obstruction than a 16 FPI rad. They can do that, but I picked a rad that is a proxy for an average heatsink; many of them provide less obstruction than a 16 FPI rad. This rad was chosen because it presents an obstruction that is average for a large number of heatsinks.

So go back and commune with the information. It will repay you to do that.

BTW -- in you 2d paragraph, you note that the NF-A12x15 beat an NF-S12A in a case (presumably through a dust filter). That is exactly what the first chart shows.

ehume said:

These were Real World Tests (tm).

So go back and commune with the information. It will repay you to do that.

Click to expand...

thank you for the info. yeah, i agree there is a lot of info in those graphs, i jsut cant seem to understand how to read them. i may be a bit slow.... also, i dont tend to spend much time looking at fan graphs so im sure there might be a little learning curve.

i was unsure if i could post links. thats why i said to go to you tube and search that exact phrase. the first video shows the heatstink and test they had. it was semi real world and it showed that 1 a12x15 matches 2 nf-f12's. i was hoping someone with the right gear could confirm that test. i know most fans seem to fall in a margin of error of each other. but i have seen other scatter plots that at least narrow down bad fans from good ones. with these 2 specific fans im sure the plots would be similar, but i would still find it very interesting to see where/how they do in that regards.

it seems like the a12 is much better from what you are saying. isnt 3db a big difference as far as actual percepted noise? so its just as good as a f12, but half as loud? i thought i read somewhere that sound doubles every 3db?

also, like you said how the the f12 performs better in open air. thats the point that noctua makes in the video, that many fan companies test in ways that are not practical. so on spec sheet, a certain fan might have a higher CFM (when tested with 100% obstruction). but in real world on a rad, it could be worse. thats where they say the a12 shines, in real world situations, in a case, on a heatsink or on a rad.

if the a12 was 6cfm better i am curious what that means in reagards to temps? is it 1 degree better, 5 degrees or none?

carlosriosness said:

. . . those graphs, . . . how to read them.

Click to expand...

Pick your condition -- case/heatsink/rad -- then pick a chart and look at the middle stripe.

with these 2 specific fans im sure the plots would be similar, but i would still find it very interesting to see where/how they do in that regards.

Click to expand...

Indeed. They were picking a very good fan to compare it with. I believe the blade angle makes the difference. I can't prove it though. But this fan reminds me of GT's.

it seems like the a12 is much better from what you are saying. isnt 3db a big difference as far as actual percepted noise? so its just as good as a f12, but half as loud? i thought i read somewhere that sound doubles every 3db?

Click to expand...

3dB doubles the power, and it is the perceptual threshold, where the normal listener can tell that one item is louder than another. But at 10 dB something is perceived to be twice as loud as something else.

. . . open air. thats the point that noctua makes in the video, that many fan companies test in ways that are not practical. so on spec sheet, a certain fan might have a higher CFM (when tested with 100% obstruction). but in real world on a rad, it could be worse. thats where they say the a12 shines, in real world situations, in a case, on a heatsink or on a rad.

Click to expand...

You are correct, except that the only identical comparison is an unobstructed fan, so all companies use that as a spec, even Noctua. All airflow specs are impractical, which is why we test them. And the NF-A12x15 indeed shines.

if the a12 was 6cfm better i am curious what that means in regards to temps? is it 1 degree better, 5 degrees or none?

Click to expand...

It differs from rad to rad. You have to know your own equipment.

thank you for all that info. it helps!

so if i look at the 3rd chart (30fpi usage). its saying that the a12x15 is pushing 46cfm while at 27 dB? so in that case the san ace silent M pushes 55CFM at 26dB? that san ace seem to be really well balanced then, right?

Screws for Rad mounting?

How did you mount this fan to radiators without the screw hanging out 5 mm? I can't seem to find what size screws I would need and it doesn't come with them

http://www.homedepot.com/p/Klein-Tools-Triple-Taps-62532/203310531
use any 6/32 screw cut to length you want

||Console|| said:

http://www.homedepot.com/p/Klein-Tools-Triple-Taps-62532/203310531
use any 6/32 screw cut to length you want

Click to expand...

Thank you, yeah i understand that part, I was more hoping to find out what length screws these would be, 20mm I assume but I know I'm going to end up buying like 9 different sizes.

15mm + maybe 2-3 to get into the rad .
I would put one in mark with a sharpe then cut all to that length

carlosriosness said:

thank you for all that info. it helps!

so if i look at the 3rd chart (30fpi usage). its saying that the a12x15 is pushing 46cfm while at 27 dB? so in that case the san ace silent M pushes 55CFM at 26dB? that san ace seem to be really well balanced then, right?

Click to expand...

The NF-A4x15 costs $15. The San Ace non-PWM retails for $22 -- 50% more; another league. I got mine in a lot of eight when the reseller ordered too many for a customer. That is the same: "Silent" fan that Seasonic puts in the X-Series PSU's. It competes with the Gentle Typhoon. If you can get a PWM in the "Silent" series, get one. I believe the 9S1212P4M011 is the only PWM they make on a routine basis, but any of the manufacturers will make you anything if you pay enough.

@bzh2 and @||Console|| - you generally buy a screw that is 5mm longer than your fan is thick. Most rads use 4mm scews, though I have seen a 3mm screw.

ehume said:

The NF-A4x15 costs $15. The San Ace non-PWM retails for $22 -- 50% more; another league. I got mine in a lot of eight when the reseller ordered too many for a customer. That is the same: "Silent" fan that Seasonic puts in the X-Series PSU's. It competes with the Gentle Typhoon. If you can get a PWM in the "Silent" series, get one. I believe the 9S1212P4M011 is the only PWM they make on a routine basis, but any of the manufacturers will make you anything if you pay enough.

@bzh2 and @||Console|| - you generally buy a screw that is 5mm longer than your fan is thick. Most rads use 4mm scews, though I have seen a 3mm screw.

Click to expand...

it was just a rough guess

bzh2 said:

How did you mount this fan to radiators without the screw hanging out 5 mm? I can't seem to find what size screws I would need and it doesn't come with them

Click to expand...

Standard screw length for 25mm fans is 30mm. So 20mm screw for a 15mm fan sounds right.

Hello all. New to the forum.

I have ben using a NF-f12 fan on the condenser of my 12v compressor refrigerator for close to 5 years.

I replaced the original fan which was rated at around 72cfm and was sucking air through the condenser. The condenser is very similar in design to the finned heatsinks I see in the Noctua videos. the screwholes on the condenser line up with a 120MM fan.





I installed the NF-f12 to push air through the condenser and the amount of time the compressor ran instantly was reduced by anywhere from 30 to 45 seconds. from over 5 minutes to under five minutes. That and the fan is quieter, and consumes 0.07 amps less, and I have been quite happy with my modification and the Noctua fan.

This new Noctua Fan, one attached to a heatsink is claiming to outperform 2 NF-f12 in a push pull situation, yet the static pressure rating is but a fraction of the nf-f12. I get that cfm rating and SP ratings are not the end all be all of measuring performance, but what has gotten my attention, is the fan they are showing in the videos from the taipei show as Noctua newest 120 m offering in the A series, is showing a 9 blade impeller, yet the data sheets on the A12x15 shows 7 blades, unless I am missing something.

If this new fan is as capable as two push pull nf-f12's at only 15 mm thick, well that extra 10MM is 10 more mm of open space behind my fan for it to scavenge air from, currently about 45mm clearance until the cabinet wall in which is resides.

I could easily retask my current Nf-f12 and have been meaning to get another one to help exhaust a cabinet into which the heater condenser air blows and also houses Inverters and solar charge controllers. This New fan would seem to be a no brainer, but I cant get over the fan they are showing in the video having 9 wider blades. In fact I see no Noctua offering in the 120 or 140Mm size which has 9 blades.

Is is possible this single fan outperforming the push pull dual nf-f12s is not the a12x15?

The one fan shown in the beginning of " Over four years in the making, Notcua’s A Series fans are finally here" video on youtube, looks like no Noctua fan I can find in their lineup. The Impeller looks more like a gentile typhoon.

What Am I missing?

In the review there is a clear shot of the 12x15 fan -- 7 blades. It puts out 46 CFM on the 30FPI rad, compared with the F12's 40 CFM.

If you are satisfied with the performance of the F12, stick with that. It is under warranty for another year, and should last longer than that. When it comes time to replace it, though, you might consider the 12x15 -- or they might have something better by then.

Thank you for your in depth reviews eHume

I was confused by the youtube videos showing a Noctua fan that appears to be 120MM and has 9 blades on the impeller.



Any Idea which fan that is, or is is some unavailable prototype?

I though perhaps they were running a ringer in their dual nf-f12 vs the single a12x15 fan test in in the video.

I am very satisfied by the nf-f12 on my fridge condenser. The sleeve bearing fan the fridge came with has been employed elsewhere, and just recently decided it will not start on its own without a manual nudge of the blades.

I was considering getting another Nf-F12 to install in another portion of my cabinet for the low noise and low amp draw. There currently resides an Addo 80mm fan on a noctua LNA cable to slow it sown to tolerable noise levels.

This fan will draw some air through the fridge condenser depending on how I have the partition set up.

Not sure the fins per inch of my condenser, but your tests show the A12x15 outperforming the nf-f12 then perhaps I can shave more time of the compressor run times and use less battery power.

The fan on the condenser only comes on with the fridge compressor the other cabinet fan runs 24/7.

I thought the a12x15, being thinner might allow more room near condenser to pull air, and also be able to push more air through condenser. The extra 0.08 amps is not really a concern, especially if it can further reduce compressor run times. I could then take the older NF-f12 and install it where the despised 80mm Addo fan resides and cool other electronics better and also work in conjunction with the fan on the condenser as the cabinet is pretty tight.

I keep havnig issues with no cursor presenting itself in the reply box. This reply was delayed several hours because of it. Is it a new member delay?

I know nothing about any delay.

The fan you saw is the prototype 12x25mm fan, like your NF-F12. However, it looks like a Gentle Typhoon. Noctua's PR guy says the hub is shaped that way because of the bearing -- all industrial fans look that way. So their 12x25mm will have 9 blades, but the current A12x15 has seven.

Noctua also makes 80mm fans, for a drop-in replacement for your Adda.

The review also shows the 12x15 equaling the F12 on the 16FPI CFM, but 3dB quieter.

But for a cabinet fan I would recommend the NF-P12, but they don't make that anymore. Or I would recommend a San Ace Silent type -- 9S1212x40y, where x = S, M, F, H and y = 1,2; available from Newark.com (you will have to put on your own cable ends) -- or the Gentle Typhoon -- or waiting for the prototype Noctua GT-clone when it goes into production.

NB -- The Noctua guy says they developed the prototype fan on its own, but I note the GT has been around a while; I get the impression some key GT patents have expired or will soon expire. Expect GT clones to hit the market when the patents expire.

bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb bbbbbbbbbbbbbbbbbbbbbbbbbbbbbb

I can't get a blinking cursor in dialog box in Chrome. Fired up firefox first time in a long time just to respond. sorry about the bbbb, don;t even know how that posted.


The a12x15 should open up 10 more MM of space behind my fan, though I am not sure the existing restriction is impeding flow on the NF-f12.

If I acquire it, then I have the existing NF-f12 to move to where the unacceptable 80mm Addo resides. Opening up the thin plywood to accommodate 120mm is not an issue. This is just a plywood cabinet in a caravan. The 80mm Addo fan currently pulls some air through condenser when main cabinet door is closed. I have it slowed way down with Noctua's LNA cable so I cannot hear it over the other fans usually running. Below it resides solar charge controller, inverters and chargers.

Out of curiosity why the P12 over the F12 for the cabinet fan? Is it better with restriction behind rather than in front of the impeller?

Thank you for the clarification of the Noctua prototype with 9 bladed GT like impeller.

My ceiling vent currently employs a silverstone fm121 and attached to the SS fan is an Arctic f12 open case fan which counterrotates in relation to SS fm121, and acts as a feeder fan for the silverstone. These together blow through a 4 inch step down ring and into a mushroom vent There is considerable restriction in front of the fans. The SS fan's noise at max speed is greatly lessened by switching on the arctic F12 fan, and Airflow also appears to be much higher.

Anybody else consider feeding a fan with another counterrotating fan? I tried two fans inline 10MM apart or so, which spin the same and the noise usually increases and airflow decreases, but with the counter rotating fan, noise decreases and airflow increases.

I want to replace the SS fm121 with a Noctua 3000 rpm industrial 120mm and their speed controller to make this roof exhaust even more effective. Will likely employ the F12 as a feeder fan again as it is already there. The Noctua Ipppc3000 should be much better at forcing air through the step down ring/ mushroom vent, and use less battery power when doing so. In theory.