Airflow vs static pressure fans. Are they a scam?

[claes]

Great posts @Valantar and @bug

I’ve been searching through the SPCR archives and ehume’s work to find a concise article, but no luck now that those sites are dead

I do want to add that noctua had been pretty candid about all this. I know they leave a sour taste in a lot of peoples mouth’s, but they have some useful values (check out those PQ curves — they have more docs like this but it’s redundant outside of this piece).

NF-A12x25: interview with Lars Strömbäck (Noctua CTO)

Designed in Austria, Noctua's premium cooling components are renowned for their superb quietness, exceptional performance and thoroughgoing quality.

noctua.at

 

[Ominence]

It's no for you, but for me it works well. Intake fans help a lot too cool GPU, while keeping system quiet. And dust for me somehow doesn't really accumulate fast, so I never care about minimizing it with fan setup.

either you are in a very low dust environment, have no filters on the intake side or your intake fans are doing surprising less than you might imagine.

 

[The red spirit]

It's also wrong. But to hell with the countless engineers that have been working on these for ages, surely they haven't thought about computing airflow volume.

I'm not shitting at engineers, but that formula in Sanyo website is only good for making airflow-static pressure curves. That site also discloses "a typical fan", but what if you increase diameter or thickness? It doesn't account for that.

either you are in a very low dust environment, have no filters on the intake side or your intake fans are doing surprising less than you might imagine.

I wouldn't call environment dust free or low dust, but I use two Silencio FP fans at 1080 rpms as intakes. That's not much of airflow, but I don't need much anyway. And from my testing, it seems that increasing intake fan speed has a very direct correlation with GPU temps and fan rpms.

 

[doyll]

Computer fan specification are mostly goobledygook.

Airflow rating is maximum airflow with not restrictions at maximum speed. Fan test stations have 2 chambers with fan being tested mounted between them. Each chamber ha it's own fan to balance pressure between chambers. Some testing is done with fan being tested have equal pressure both intake and exhaust side, a condition a fan in actual use can never have. Airflow is generated when higher pressure air move into lower pressure are trying to make both equal in pressure. Our fans draw air from intake side (creating lower pressure area) on intake side and push air out their exhaust side (creating a higher pressure area) on exhaust side.

So our fan specs are the result of unnatural conditions .. nothing like the conditions when installed on a cooler / radiator or in a case.

But let us say airflow rating is how much air fans can move with no airflow restrictions like grill, filler cables, etc .. like a fan mounted on a stand blowing air in an open room.
And static pressure rating is how much pressure a fan can push into a sealed box. For the fun of it let's say this fan pushes 1.836 mm H2O of pressure into our box.

1.386 mm H2O static pressure rating, the difference between intake and exhaust sides is extremely low .. so low it's almost impossible to measure.
The barometric pressure at sea level and 30 feet above sea level is 11.013 mm H20.

Barometric pressure is the force (pressure) air is pushing against our body.​

That is a difference between sea level and 10 feet above sea level of 3.671 mm H2O.
Which is a difference between sea level and 5 feet above sea level of 1.836 mm H2O
Few of our fans have 1.836 mm H2O static pressure rating Most have much lower pressure ratings.
1.836 mm H2O pressure rating is the difference in pressure between the air pressure on our feet versus the air pressure on our chest when standing at sea level.


Now let's look at typical airflow resistance in our computer systems.
Grills lower airflow 29% to 71%. Open link below and scowl down to see test results.

SilverStone

Founded in 2003, SilverStone is an established leader in its field, with an elite team of engineers; we started our quest of providing products that create inspirations. We have since expanded the lines of products as well as types of products we produce, giving our customers a wide selection...

www.silverstonetek.com

Filters add even more airflow restriction.
Limited space on intake and exhaust sides of fans also restrict airflow.
Basic rule of thumb is fans should have their diameter in clearance on both sides. This means a 120mm fan should have 120mm of free space on intake and exhaust side.
Only vent in most cases with this clearance is rear exhaust and maybe top.
At a bare minimum we should have 1/3rd the diameter in clearance, which is 40mm on each side.
More cases have this kind of clearance.

I could go on and on and on about how specs are at maximum speed and we run our fans at maybe half speed where pressure and airflow ratings are less than have of spec. But this is already a long post.

My advice is choose fans with high pressure ratings paying no attention to airflow rating.
Biggest challenge our fans have is overcoming airflow resistance.

 

[claes]

OMG is that @doyll from OCN? Long time no see lol

 

[Ominence]

I wouldn't call environment dust free or low dust, but I use two Silencio FP fans at 1080 rpms as intakes. That's not much of airflow, but I don't need much anyway. And from my testing, it seems that increasing intake fan speed has a very direct correlation with GPU temps and fan rpms.

GPU temps for a given noise level perhaps benefit with intake fans just across them. i like to overcome that issue with brute force exhaust. different approach i guess.

i like to overcome that issue with brute force exhaust.

by that i mean my case fans are linked to the GPU temp probe.

 

[The red spirit]

GPU temps for a given noise and temp level perhaps benefit with intake fans just across them. i like to overcome that issue with brute force exhaust. different approach i guess.

Well, you can always overcome issues with lots of rpms, but I don't want noise.

 

[claes]

@The red spirit your case (meaning your chassis) is pretty unique/totally normal in this regard, in that it’s not representative of most fan test environments. It makes a lot of sense that negative pressure would work out well for you — might even consider removing your PCI brackets for fun

 

[Ominence]

Well, you can always overcome issues with lots of rpms, but I don't want noise.

with 15000rpms of case fans mine is an absolute jet at cold post. very quite from there on depending on the fan curve. 20% duty cycle at idle or 3000rpms combined. my aircon is louder than the pc.

 

[Valantar]

I don't see any reason why they sometimes couldn't be so. I remember reading that there were 40mm pc fans created, but it was a very niche product, which wasn't advertised much and most people didn't even know that it existed, however, few that have those fans claim that they are great. And for example on tower cooler like 212 Evo, it wouldn't be too difficult to mount 40mm thick fan and it would be possible to have cooler more dense than it already is, so that it performs better. In most cases, you can mount 40mm fans as exhausts too. Intakes are thought due to various obstructions, but I think that 40mm fans certainly could be sold for pc enthusiasts wanting to try something different. And if Corsair, CM or Noctua would release such model, it could gain become a reasonably well known.

They can't be thicker than 25mm without breaking the (de facto? I have no idea if this is official in any way) standard for PC case and cooler fans, risking severe compatibility issues. There are many places where a >25mm fan can fit, but that doesn't mean that producing such a fan would be unproblematic. With all else being equal, the return rate for a thicker fan due to it not fitting the intended use would be higher simply due to it being thicker than standard.

Of course designing a thicker fan for consumer use is also problematic in that no such fans exist, and adapting a server or industrial design is likely to be difficult (different factors to consider). In other words: it would be expensive. And as I said, fans are a commodity, and costs matter.

Then there's the actual benefits: unless you are running a very thick and restrictive radiator, current 25mm fan designs perform pretty well, and significantly beating that without getting to the range where two 25mm fans in push-pull would be an alternative is very difficult, especially at low noise levels. And as pressure stacks with fans in series, two fans of X mmH2O = 2X mmH2O (minus some interference losses). Thus, a 40mm+ fan would need to outperform not one, but two 25mm fans stacked, while still staying at a reasonable development cost.

There are some thicker fans out there, like EK's 38mm Meltemi fans, though those only really outperform 25mm fans in specific scenarios (in particular: restrictive rads in pull configuration).

lol it was me who said that they are not like Archimedes screw (which I think would be reasonable static pressure fan design, but who knows, if it spins too slow, water can just go back and air would be even harder to start moving. Anyway, since you say that computers will always be restrictive (which is definitely true, even more so for high end systems), wouldn't that mean that you would always want to buy only static pressure optimized fans? Because as you said too, there isn't much from where airflow fan would have unimpeded airflow and thus won't performs nearly as well as they are stated in spec sheet.

As you say, an archimedes screw would be a pretty terrible air pump though - you need something that seals, i.e. a piston mechanism or similar for that. But you seem not to have read very carefully: I also point out that all PC fan designs fall along a relatively narrow spectrum of possible designs. This means that most "airflow optimised" fans still produce a modicum of static pressure when restricted, and that most "pressure optimized" fans still provide good airflow when unrestricted. So while there are differences, they are small and nuanced. If your case is extremely unrestrictive and has a good airflow layout with equal intake and exhaust, a good set of airflow optimized fans will likely provide more airflow through the case at the same noise level as a set of good pressure optimized fans. But you'd still want a more pressure-optimized fan attached to any heatsinks in the case, as airflow will otherwise choose the path of least resistance and any airflow-optimized fans mounted to these will perform somewhat worse. Not a world of difference, but some (as long as all fans are assumed to be good designs). But if, for example, your case has less affordances for air to leave the case than enter it (forcing a positive pressure setup), then yes, you want fans better optimized for pressure, as you'd otherwise risk losing effective airflow due to back-pressure. Again, the difference isn't likely to be major, but it would be measurable.

Edit: in a case like yours, which is highly restrictive by design, airflow will always be hampered regardless of your choice of fan, but good "pressure optimized" fans are still likely to perform better than "airflow optimized" ones.

 

[Ominence]

They can't be thicker than 25mm without breaking the (de facto? I have no idea if this is official in any way) standard for PC case and cooler fans, risking severe compatibility issues. There are many places where a >25mm fan can fit, but that doesn't mean that producing such a fan would be unproblematic. With all else being equal, the return rate for a thicker fan due to it not fitting the intended use would be higher simply due to it being thicker than standard.

how does a 125mm stacker 15000rpm 'sound' with my setup. 160mm and 18000 peak rpm if you include the cpu fan blowing very close in the same direction, and the single case grill in the middle. i have a stupid setup, thought it nice to give people a good laugh and cringe for others.

 

[Valantar]

from a low maintenance perspective, i've to say no to wasting any fans on the intake side. the moment your airway path immediately before or immediately after (filters or grills) the fan blade is restricted, the cfm of HSP will still be higher than that of the so called HAF fans.

You have no intake fans at all? That's a balancing act, and highly dependent on the case. The (negative) pressure produced by the exhaust fans then have to work on a much larger area (every available opening in the case), so depending on the size of the case and how restrictive it is, as well as how well positioned these air intakes are to components in need of fresh air, it can either be really great or downright terrible. A lot of sandwich layout SFF cases with radiator support perform drastically better with all fans set as exhaust, pulling fresh air in through ventilated side panels. But in other cases the fresh air might bypass the components in need of cooling entirely - it depends where it can get in more easily, and as pressure will decrease in the case the further from the fan you are (unless your case is a near-perfect wind tunnel with equal active exhausts and passive intakes), air will be more likely to enter closer to the fan if it is able to.

how does a 125mm stacker 15000rpm 'sound' with my setup. 160mm and 18000 peak rpm if you include the cpu fan blowing very close in the same direction, and the single case grill in the middle. i have a stupid setup, thought it nice to give people a good laugh and cringe for others.

15000 rpm? Unless you're running some sort of high speed server fans, you can't add up rpms. Stacked fans add pressure, not airflow (except when said pressure serves to overcome a restriction, which is always up to a cut-off point), and rpm is effectively the same no matter the amount of fans. Of course stacked (serial) and side-by-side (parallel) fans affect airflow in a system in very different ways - side-by-side fans do add up in airflow, but not really in pressure (that Sanyo link on the previous page illustrates this nicely). Parallel fans increase swept area, stacked fans increase pressure.

 

[The red spirit]

@The red spirit your case (meaning your chassis) is pretty unique/totally normal in this regard, in that it’s not representative of most fan test environments. It makes a lot of sense that negative pressure would work out well for you — might even consider removing your PCI brackets for fun

How it is even a bit unique? I only have Silencio S400, a small case, packed with lots of impedances. It's pretty much looks like any other generic mATX modern silent case. Define C Mini should perform similarly to this S400. Anyway, I have two stock fans (Silencio FP 120mm) set at 1080 rpm as two from intakes and due to DVD drive restriction, I use two exhaust fans, which are both Silverstone Globe 120mm rpms fans (from Grandia GD05B), one is rear exhaust and other is rear top exhaust. They both run at less 650-680 rpms. That's slow, but I noticed that they get a lot noisier than Silencio FPs at same rpms and so my current setup is setup more for minimal noise, rather than ideal thermal performance. There thermal performance of my machine is satisfactory, meaning it's quiet enough so I don't care and cool enough, so it stays quiet enough and safe enough. If I cared more about airflow, I would replace Globe fans with same Silencio FPs as they seem to be exceptionally quiet at quite a bit more rpm. But since I don't have any particularly hot hardware, there's not much reason for me to mess with fan setup. In fact I could remove two fan and just use those two Silencios only, but then my graphics card runs louder and I don't like that. With current setup graphics card runs Unigine Heaven benchmark and stays at around 1300 rpm. And I tolerate only under 1500 rpm. And now are one of the hottest summer days too.

And it would seem that I have positive pressure setup, but that's most likely not a case either. Those Silencio FPs have high pressure rating and a lot lower cfm rating, so they probably in real life also push less real cfms too and those two Silverstone Globe's at less than 700 rpm likely match them in terms of airflow, since their only restrictions are case grille an top fan filter. In terms of dust, I have been using this case for like 3 months and front filter is a bit dusty. Case is a bit dusty, but heatsinks are pretty much perfectly clean. Only PSU filter had some serious dust, but that's it. I don't remember having system this decently clean in Cooler Master K280 case, so S400 is an improvement.

 

[Valantar]

And it would seem that I have positive pressure setup, but that's most likely not a case either. Those Silencio FPs have high pressure rating and a lot lower cfm rating, so they probably in real life also push less real cfms too and those two Silverstone Globe's at less than 700 rpm likely match them in terms of airflow, since their only restrictions are case grille an top fan filter.

Remember that your front fans are severely impeded by the solid front panel and the narrow front-side air intakes. Airflow loses ~50% of its pressure over a 90° turn, and there's a filter there too, so they are likely quite restricted. Unless you're noticing dust ingress along case seams and cracks you're likely at rough equilibrium or slight positive pressure though.

 

[The red spirit]

They can't be thicker than 25mm without breaking the (de facto? I have no idea if this is official in any way) standard for PC case and cooler fans, risking severe compatibility issues. There are many places where a >25mm fan can fit, but that doesn't mean that producing such a fan would be unproblematic. With all else being equal, the return rate for a thicker fan due to it not fitting the intended use would be higher simply due to it being thicker than standard.

Well, that's marketing's job to inform user that these are thicc bois and that you can't just simply buy and mount them anywhere. Seems pretty easy to to pull off. But if you are paranoid, you can just ship them with CPU cooler, so that people would pay attention to dimensions.

Of course designing a thicker fan for consumer use is also problematic in that no such fans exist, and adapting a server or industrial design is likely to be difficult (different factors to consider). In other words: it would be expensive. And as I said, fans are a commodity, and costs matter.

Still, I see no reason why computer luxury brands like Corsair or Noctua couldn't experiment with that. Their buyers aren't always geniuses and they sure are willing to buy some poor value products, because Noctua, because Corsair. And I said that somebody already made those and sold them. I really can't recall who, but it may be Delta or other OEM, who specialize mostly in servers.

I also point out that all PC fan designs fall along a relatively narrow spectrum of possible designs. This means that most "airflow optimised" fans still produce a modicum of static pressure when restricted, and that most "pressure optimized" fans still provide good airflow when unrestricted. So while there are differences, they are small and nuanced.

So it's a better idea to err on static pressure side, if you have no idea how restrictive your case insides and grilles are?

Edit: in a case like yours, which is highly restrictive by design, airflow will always be hampered regardless of your choice of fan, but good "pressure optimized" fans are still likely to perform better than "airflow optimized" ones.

That's why it came with two Silencio FP fans.

 

[Ominence]

You have no intake fans at all? That's a balancing act, and highly dependent on the case. The (negative) pressure produced by the exhaust fans then have to work on a much larger area (every available opening in the case), so depending on the size of the case and how restrictive it is, as well as how well positioned these air intakes are to components in need of fresh air, it can either be really great or downright terrible. A lot of sandwich layout SFF cases with radiator support perform drastically better with all fans set as exhaust, pulling fresh air in through ventilated side panels. But in other cases the fresh air might bypass the components in need of cooling entirely - it depends where it can get in more easily, and as pressure will decrease in the case the further from the fan you are (unless your case is a near-perfect wind tunnel with equal active exhausts and passive intakes), air will be more likely to enter closer to the fan if it is able to.

yup, no intakes- all unobstructed openings with the stock fan filters everywhere else. FDDR6 case- no front door or top closed panel and with all exhaust fans on the back fan location. the 2nd cpu fan is the turbulence generator for a lot of fun airpaths within the case. that 2nd fan blows towards the front of the case! and i guess the psu fan must do a little too with the fan facing up and exhausting out the back as well.

 

[The red spirit]

Remember that your front fans are severely impeded by the solid front panel and the narrow front-side air intakes. Airflow loses ~50% of its pressure over a 90° turn, and there's a filter there too, so they are likely quite restricted. Unless you're noticing dust ingress along case seams and cracks you're likely at rough equilibrium or slight positive pressure though.

I'm not so sure about exactly losing a lot of pressure, because opening doors does surprisingly next to nothing. As for grille, it really doesn't look very restrictive. So in reality, I probably don't lose much pressure. Still, Silencio FPs from hand measurement don't provide much airflow so even without pressure loss, in my case there's probably only a tiny bit positive pressure or neutral pressure.

 

[Ominence]

15000 rpm? Unless you're running some sort of high speed server fans, you can't add up rpms. Stacked fans add pressure, not airflow (except when said pressure serves to overcome a restriction, which is always up to a cut-off point), and rpm is effectively the same no matter the amount of fans. Of course stacked (serial) and side-by-side (parallel) fans affect airflow in a system in very different ways - side-by-side fans do add up in airflow, but not really in pressure (that Sanyo link on the previous page illustrates this nicely). Parallel fans increase swept area, stacked fans increase pressure.

i'm kidding- that's 5x3000 stacked together. i don't think noobs will be too interested in this thread so surely no one should really believe it, like you didn't! i do not intend any misdirection- just fun participation. i would love to do an rpm test one day on the last serial fan in the stack. that should be doing some 'fun' rpms regardless of the 3k rated speed. force fed induction is great.

 

[80251]

I think most consumer grade fans are a joke. SP vs AF? Server fans have both. Common misperception is that they are loud.. yes.. yes they are. But only at full speed. What is SP? It’s just air.. but more of it. Thick fans are where it’s at.

There are some amazingly thick and powerful fans made by Delta, Nidec and SanAce. Most of them are vaneaxials, some of them are contra-rotating dual fan assemblies. The amount of static pressure and CFM these fans put out makes anything by noctua look like a joke.

 

[qubit]

Static pressure = pressure. Just a different term for the same thing.

 

[The red spirit]

There are some amazingly thick and powerful fans made by Delta, Nidec and SanAce. Most of them are vaneaxials, some of them are contra-rotating dual fan assemblies. The amount of static pressure and CFM these fans put out makes anything by noctua look like a joke.

Joke? Sorry mate, look at specs. Here is Noctua NF-A12x25 vs Delta AFB1212L-R00:
rpm 2000 vs 1900
cfm 60.09 vs 62.93
mmH2O 2.34 vs 3.93
dBA 22.6 vs 32.5
MTTF 150k vs 70k

While air moving specs are great, Delta fan is a total disaster in terms of noise. Also Delta fan may or may not last shorter. Meanwhile, Noctua's Industrial NF-F12 is better than Delta at every single measure. It moves more air, it has more pressure, it lasts longer is quieter and consumes less power.

And what you could have expected, Delta just makes generic 7 bladers, doesn't give any dampening rubber, uses ball bearing. It's surprising that it wasn't beaten any worse than it already is.

Nidec doesn't even make anything that doesn't rape your ears and something that doesn't set motherboards fan connector on fire. And they sell some crazy stuff. 9800 rpm fan? They make that:

V12C-M3 | NIDEC CORPORATION

www.nidec.com

They also don't disclose what connector their fans have.

Sanyo Denkis are also no match for Noctua Industrials and on top of that they fail to deliver as much pressure as 7 blader Delta. So they are boring too. And they don't make any crazy near 10k rpm fans either, super boring. And yet they still burn motherboards.

 

[80251]

Wow the ignorance of the type of fans Delta, SanAce and Nidec manufacture is remarkable. Delta, Nidec and SanAce are renowned for their longevity because they're designed to work in server environments -- 24-7-365. I've never come across anyone working in any datacenter who uses any Noctua fans whatsoever -- because they're far too weak. No server manufacturer uses Noctua fans either.
Noctua makes very few fans that have more than 100CFM. Why don't you compare fans that produce > 100 CFM? Noctua doesn't even have ANY fans that produce > 150CFM and I don't trust the specs on their NF-A14 iPPC-3000.

 

[Valantar]

Well, that's marketing's job to inform user that these are thicc bois and that you can't just simply buy and mount them anywhere. Seems pretty easy to to pull off. But if you are paranoid, you can just ship them with CPU cooler, so that people would pay attention to dimensions.

You can't inform consumers out of assuming a product that breaks an ubiquitous standard is standard-compliant. Most buyers don't look that closely. There will always be a higher return rate for a product like this, and marketing can't do anything to fix that.

Still, I see no reason why computer luxury brands like Corsair or Noctua couldn't experiment with that. Their buyers aren't always geniuses and they sure are willing to buy some poor value products, because Noctua, because Corsair. And I said that somebody already made those and sold them. I really can't recall who, but it may be Delta or other OEM, who specialize mostly in servers.

They have no doubt done the research and concluded that it's not worth it. For heatsinks, a thicker fan means less fin area for the largest ones - space in a case is finite, after all - where the better fan would be of the most benefit. And a thinner fin stack is less restrictive, meaning the thicker fan would lose its main benefit. For case fans, you can't generally assume that all cases are wide open around their fan mounts. And again, it might not provide a sufficient benefit to actual airflow without sacrificing other metrics. Also, I think you're being a bit too optimistic about how easy it is to amortize the R&D cost of a brand new, non-standard fan design across sales, especially when sales will be lower due to compatibility issues.

Also, I mentioned one of the more available thicker options above, the EK Meltemi 38mm. It's not particularly good - it outperforms the 25mm Vardar on thick radiators in pull configurations, but that's about it. In push or with a thinner radiator it's no better.

So it's a better idea to err on static pressure side, if you have no idea how restrictive your case insides and grilles are?

Yep, pretty much. But as I said, differences between well designed fans are relatively small, and most "AF" fans are much closer to SP fans than actual AF fan designs.

i'm kidding- that's 5x3000 stacked together. i don't think noobs will be too interested in this thread so surely no one should really believe it, like you didn't! i do not intend any misdirection- just fun participation. i would love to do an rpm test one day on the last serial fan in the stack. that should be doing some 'fun' rpms regardless of the 3k rated speed. force fed induction is great.

Sheesh, that sounds ... interesting I kind of doubt that last fan will be spinning much over spec though - stacking fans only increases pressure and serves to overcome restriction, flow stays the same (or might even drop due to turbulence between the rotors). It will likely be spinning faster than if it was alone simply due to more pressure allowing it to operate with less restriction, but the difference won't be huge.

Wow the ignorance of the type of fans Delta, SanAce and Nidec manufacture is remarkable. Delta, Nidec and SanAce are renowned for their longevity because they're designed to work in server environments -- 24-7-365. I've never come across anyone working in any datacenter who uses any Noctua fans whatsoever -- because they're far too weak. No server manufacturer uses Noctua fans either.
Noctua makes very few fans that have more than 100CFM. Why don't you compare fans that produce > 100 CFM? Noctua doesn't even have ANY fans that produce > 150CFM and I don't trust the specs on their NF-A14 iPPC-3000.

Noctua doesn't target servers, as their niche is silence, and nobody cares about silence in a server. So ... yeah. That point is kind of moot. You're not getting a silent > 150CFM 120mm fan either, no matter what you do. As for reliability, I've never heard of Noctuas not handling 24/7 running (what's with people operating with 365-week years? Mine only have 52.), but obviously they're not 10 000rpm server screamers, and couldn't handle those kinds of loads. They aren't made for that. Different tools for different purposes.

Also, most server fans have dual ball bearings, which makes them grindy and noisy at low to medium RPMs. Just ask any Gentle Typhoon owner who has since experienced a good non-ball-bearing fan. The difference is very noticeable.

 

[The red spirit]

Wow the ignorance of the type of fans Delta, SanAce and Nidec manufacture is remarkable. Delta, Nidec and SanAce are renowned for their longevity because they're designed to work in server environments -- 24-7-365. I've never come across anyone working in any datacenter who uses any Noctua fans whatsoever -- because they're far too weak. No server manufacturer uses Noctua fans either.

I'm pretty sure that they just don't wanna pay for luxury fans.

Noctua makes very few fans that have more than 100CFM. Why don't you compare fans that produce > 100 CFM? Noctua doesn't even have ANY fans that produce > 150CFM and I don't trust the specs on their NF-A14 iPPC-3000.

Because those high CFM fans spin at crazy rpms and are loud, if not outright burning every motherboards' fan connectors. You know, they have maximum power ratings too and they surely aren't made to handle 20+ watt fans.

You can't inform consumers out of assuming a product that breaks an ubiquitous standard is standard-compliant. Most buyers don't look that closely. There will always be a higher return rate for a product like this, and marketing can't do anything to fix that.

I doubt that. I meant the whole mITX market exists and they are are always thinking of ways to makes computers smaller and people buy that stuff.

For case fans, you can't generally assume that all cases are wide open around their fan mounts.

I clearly mentioned that thicker fans would be for those that do their own research and I don't expect them to fit everywhere.

Also, I think you're being a bit too optimistic about how easy it is to amortize the R&D cost of a brand new, non-standard fan design across sales, especially when sales will be lower due to compatibility issues.

How hard could it possibly be. Just give review units to LTT, JayZ, Bitwit and write in big bold letters, that these fan's won't fit everywhere. Done. And write that on your website and fan box. Now, truly done. You know, it's not like we didn't have 200mm fans or even bigger 240mm fans in computers.

Also, I mentioned one of the more available thicker options above, the EK Meltemi 38mm. It's not particularly good - it outperforms the 25mm Vardar on thick radiators in pull configurations, but that's about it. In push or with a thinner radiator it's no better.

That's because it's a pressure design, give that extra thickness to airflow design and it may do wonders.

 

[DeathtoGnomes]

Different tools for different purposes.

This pretty well sums up a lot., the right fan for the right job.

Appreciate this entire discussion on fans, and vote to put in the hall of fame .