ARCTIC P12 PWM PST 120 mm Fan

[TechBuyingHavoc]

The P14s are the same, but relatively better - and by that I mean that they're exactly what you'd expect but the landscape for 140mm fans simply has fewer decent 140mm offerings.

Other sites have validated the P14 and there are zero surprises - as long as you understand the strengths and weaknesses of larger fans in general (gross oversimplification, but "more airflow at any given rpm/noise level, but worse static pressure so not as good at filtered intakes or restrictive heatsinks/radiators" covers the basics).


They probably updated their PWM chip after the P12 Max was manufactured. The IC is tiny and cheap and sourcing the old ones if the IC manufacturer updated their IC to include zero-RPM would be more hassle than it's worth - and why skip on the free upgrade? These ICs for PWM motors are a couple hundred bucks for reels of 5000, so they're fractions of a cent.

Since you know your fans, is there any reason to get the CO versions of these P12/P14 PWM/PST fans? I understand that they have ball bearings instead of fluid-impregnated sleeve bearings and should last longer. I do keep my computer on at night so it is running 24/7 at low clocks usually when not gaming on it.

Or do I just get a new fan when the regular Arctic fans wear out in how many years?

 

[Wasteland]

Since you know your fans, is there any reason to get the CO versions of these P12/P14 PWM/PST fans? I understand that they have ball bearings instead of fluid-impregnated sleeve bearings and should last longer. I do keep my computer on at night so it is running 24/7 at low clocks usually when not gaming on it.

Or do I just get a new fan when the regular Arctic fans wear out in how many years?

I'm not Chrispy obviously, but FWIW I'd avoid the CO fans. Arctic itself seems to have reconsidered their usefulness--initially all of the P12 Max had the "CO" ball bearings, but now they don't, presumably due to user feedback about the sound profile. IIRC, the timeline went something like this: Arctic released the P12 Max in black, with ball bearings, then they released the white variant, this one using the default rifle bearing (what they call "fluid-dynamic"); then they revised the black version to the rifle bearing too.

The ball bearing variant pays for its theoretical longevity advantage with extra noise. Since the normal sleeve/fluid-dynamic bearing models should last you a decade anyway, this is a bad trade off. Of course if you plan to deploy the fans in an unusually hot environment, the CO variant might appeal to you, but otherwise, you can't go wrong with the default.

 

[Chrispy_]

Yeah, ball bearings are better suited to clean environments and higher-rpm fans. An air-conditioned, dehumidified server room or datacenter rack for example.

Historically, ball bearings do not outlast good quality FDB/rifle bearings in typical consumer use cases. I don't know the exact reason for this but it's probably a combination of dust, humidity, and cycling the operating temperatures with all the start-stop cycles of typical consumer use (especially with a zero-RPM PWM control option).

Honestly, the regular P12/P14 aren't particularly outstanding fans, they're simply decent fans at a ludicriously low price. As far as I'm aware there are no 5-packs for the CO variants, so you'd be paying 3x the price per fan which makes the CO variants decent performance at a slightly higher than average price, i.e. they're okay but not worthy of any particular recommendation. Let's face it, nobody would be giving the regular P12/P14 a second glance if they cost 3x what they currently do.

 

[Knight47]

The variants with the ring around the fan blades have evidently cured the high-pitched sound-profile issue at lower RPM. These models are the MAX and the A-RGB, which are more expensive than the standard P12/P14, but still pretty hard to beat. A few links:

I recently upgraded from F14 PWM PST CO to the P14 Argb and honestly these are not as quiet as the F14.

Since you know your fans, is there any reason to get the CO versions of these P12/P14 PWM/PST fans? I understand that they have ball bearings instead of fluid-impregnated sleeve bearings and should last longer. I do keep my computer on at night so it is running 24/7 at low clocks usually when not gaming on it.

My PC runs for around 8 hours/day and the ball bearings ran dry out after 2 years in my F14 fans and they make friction noise or start to chirp. I did try 10w40 and sewing machine oil, but the bearings are closed, so no oil went it.

 

[Waldorf]

the only problem is, getting stuff from multi packs, which are prone to "failing" within 1-2y, less chance with the COs, since they go single retail pack.
in 20y of using arctic stuff, mainly their case fans, i can tell that all those i received from arctic to replace noisy ones (clicking at low speed or grinding sound from dirty bearing),
had the same issue within 2y, and all where from multi packs.

first couple years i thought it was the environment (dust), but having to regularly replace F/P-fans in the past 10y, seems to be accelerated when fans are mounted horizontal (bottom/top of case).

i mean, sure, most got replaced by arctic, but it started to get annoying with support, and regularly exchanging 2-4 fans/rerouting cables, im getting tired of using them.
might get something different, but with priority on low noise (pc inaudible at low loads), its hard to find something perf decently (flow/pressure) without costing 30$/p,
as im looking at 11 fans rn (case+rad).

 

[stahlhart]

Have the RGB version of these in all three of my current builds, including swap out of the stock fans on AIOs (22 total). Inexpensive, running 24/7 and have yet to have one fail in close to four years now.

 

[Wasteland]

the only problem is, getting stuff from multi packs, which are prone to "failing" within 1-2y

My one major beef with Arctic is that the packaging for their 5-packs is extremely flimsy. Shipping damage, obvious or subtle, isn't uncommon. Of course, it makes perfect sense for Arctic to cut costs at the price points we're discussing, but still I think they do themselves and their customers a disservice. Investing an extra couple nickels on the box would go a long way.

Anyway, that sums up my experience: if the fans aren't damaged when you receive them, they will run happily for many many years--but you do have to inspect them carefully, particularly if they came in the 5-pack. You can reduce risk here by ordering your 5-pack of fans along with other items, so that they're packed together in a bigger box.

Honestly, the regular P12/P14 aren't particularly outstanding fans, they're simply decent fans at a ludicriously low price. As far as I'm aware there are no 5-packs for the CO variants, so you'd be paying 3x the price per fan which makes the CO variants decent performance at a slightly higher than average price, i.e. they're okay but not worthy of any particular recommendation. Let's face it, nobody would be giving the regular P12/P14 a second glance if they cost 3x what they currently do.

At first I thought I disagreed on the first statement, but it's a matter of perspective. By virtue of their unbeatable price/perf, Arctic fans have become sort of a de facto standard. This makes them boring in the same way that vanilla ice cream is widely considered boring. But vanilla ice cream is delicious. It's the standard because you can't go wrong. Most people like it just fine.

Purely in terms of performance metrics, Arctic fans compete favorably with even the most expensive and over-engineered items on the market. At reasonable noise levels, they'll give you 80-90% of the benefit of products that cost 7-8 times more. And they outright beat many fans at every price point in between.

But you're right; there's nothing especially exciting here. Fans as a whole category aren't exciting (unless you're a weirdo like me). Most people just want something that will move adequate amounts of air at barely audible noise levels, when the computer is at low load. If you know how to set a fan curve, almost any fan made by a reputable brand will achieve that goal. At high loads, things can get a little more interesting, but not that much more interesting; you're still probably best off limiting your case fans to a fraction of their max speed (my Arctic fans' curve tops out at ~1200 RPM). Heatsink fans merit more interest/attention, but again most people don't run at high (CPU) load often enough to care about fan noise in that context. And if we're talking gaming rigs, the GPU's noise level tends to set the tone, and almost no one uses aftermarket fans on their GPU anyway.

Performance at high speeds is basically a marketing gimmick, at least in the desktop market. As you suggested earlier, the Arctic Max is a good product because it has excellent performance characteristics at around 30% speed--and because it's cheap and reliable--not because it can move obscene amounts of air at 3,000 RPM. It'll sound like a woodchipper at that speed. Every fan sounds like a woodchipper at that speed. This is another area where Arctic arguably did itself a disservice; the Max got a lot of hype when it released, in part because having the option to dial fan speeds up to absurd levels looks really cool on a performance chart, but when Joe Normie tosses a few Arctic Max into a machine running a default fan curve, he's likely to freak out at the noise level.

Fan tech is boring, in large part, because it's mature. There simply isn't headroom left to squeeze out substantial advantages at a remotely reasonable cost--and I do mean "remotely" reasonable; Noctua for example toyed with the idea of making the entire frame of their newest flagship fans out of liquid crystal polymer, but quickly discarded the notion due to cost. Even given that concession, the fans in question retail for ~$40. Part of me almost wishes they'd gone whole hog; it would be amusing to see how much better a $70 variant would perform, probably within margin of error.

 

[stahlhart]

My one major beef with Arctic is that the packaging for their 5-packs is extremely flimsy. Shipping damage, obvious or subtle, isn't uncommon. Of course, it makes perfect sense for Arctic to cut costs at the price points we're discussing, but still I think they do themselves and their customers a disservice. Investing an extra couple nickels on the box would go a long way.

Anyway, that sums up my experience: if the fans aren't damaged when you receive them, they will run happily for many many years--but you do have to inspect them carefully, particularly if they came in the 5-pack. You can reduce risk here by ordering your 5-pack of fans along with other items, so that they're packed together in a bigger box.

I've purchased all of mine on Amazon, and they have always been individually boxed, but it looks like for the RGB version the largest "pack" is three.

 

[Wasteland]

I've purchased all of mine on Amazon, and they have always been individually boxed, but it looks like for the RGB version the largest "pack" is three.

Yeah, my complaint is specific to the 5-pack. It's a very thin cardboard box that is just large enough to fit five naked fans, stacked together like sardines. There's zero cushion. Given the insane value on offer for the 5-pack, I don't think anyone would object to paying an extra dollar for a bit more security in shipping.

All other items from Arctic come in fairly robust packaging.

 

[Gasaraki]

I have bought high end maglev fans from Corsair over the years and 2 out of 6 of them have failed. The Artic ones I started getting recently are cheap and quiet. I think they will last longer than my Corsairs.

 

[damric]

Very good static pressure at 1800 RPM with filter on intake.

 

[Chrispy_]

Yeah, my complaint is specific to the 5-pack. It's a very thin cardboard box that is just large enough to fit five naked fans, stacked together like sardines. There's zero cushion. Given the insane value on offer for the 5-pack, I don't think anyone would object to paying an extra dollar for a bit more security in shipping.

All other items from Arctic come in fairly robust packaging.

It's Amazon, so when your 5-pack arrives damaged you send it back and Bezos personally pays for prime expedited replacements as well as your hassle-free, label-free, convenient returns at one of so many locations convenient to you.

Amazon is pretty evil so when they screw up by packaging something inadequately, they can damn well make it right.

 

[Waldorf]

@Wasteland
none were damaged during transport, or broken prior install, and working for +1-2y, so no other reason to connect this with packaging.
and im not the only one with the experience of 2 out of 5 fans going noisy. not dead in a sense of operating, but useless for use, short of above 70% rpm, while it doesnt happen on single units.

sure, there is a chance of the next 20 fans i buy to last 10y without becoming noisy, but what i had so far and know from other users, my edjucated guess is that wont happen.

i just ordered 2 with BB, see if there is a higher noise level at similar rpm, and might swap out more.
i rather have BB dry out, than clicking noise at low load.

@Chrispy_
its not amazon, its arctic, they even ship their stuff in envelopes with single layer bubble wrap, so they dont have to pay for shipping a box (shaped package).

 

[Fener]

Hi I have this case: antec p110 silent. Would you suggest that I replace the stock fans (3x120mm front and 1x140mm rear) with these? Would I benefit from more air flow and lower noise operation? thank you

 

[Waldorf]

if you are not happy with them or need to improve temps, sure, otherwise its not gonna matter much and i would rather wait until those fail and save the money.

if you want to swap, get the pXX-pwm pst co, and you can sync the top 2.

 

[Chrispy_]

Hi I have this case: antec p110 silent. Would you suggest that I replace the stock fans (3x120mm front and 1x140mm rear) with these? Would I benefit from more air flow and lower noise operation? thank you

Only buy replacement fans when you've run out of fan curve tuning options.

Moving air makes noise, if you don't like the noise and your temps are excellent, then you are paying for those needlessly low temps with noise.

Yes, some fans move air more quietly than others, but most fans are pleasantly quiet at some RPM and honestly the majority of people have their fans running too fast. Low temperatures might make your hardware last 15 years instead of 10 years, but most of the time it's replaced because it's obsolete long before that.

Nvidia GPUs target 83C so there's headroom to turn the fans down if your GPUs not even hitting high-70's. Processors will run at 85-90C all day for a decade.

 

[Waldorf]

@Chrispy_
while that works for cpu/case fans, its not useful for gpus.

last time i checked, Nv starts throttling boost clocks at 43C for every 6C increase, so the cooler it runs (under game load), the better.
one reason you wont find any aircooled gpu that boosts higher/longer than any LC version of it, unless maybe its running outside the case in a climate controlled room.

part of the reason in recommend ppl to get an AIO, so the rad dumps the cpu heat outside the case, lowering in-case temps, thus making the gpu boost higher and longer,
vs air cooled cpu.

 

[Chrispy_]

@Chrispy_
while that works for cpu/case fans, its not useful for gpus.

last time i checked, Nv starts throttling boost clocks at 43C for every 6C increase, so the cooler it runs (under game load), the better.
one reason you wont find any aircooled gpu that boosts higher/longer than any LC version of it, unless maybe its running outside the case in a climate controlled room.

part of the reason in recommend ppl to get an AIO, so the rad dumps the cpu heat outside the case, lowering in-case temps, thus making the gpu boost higher and longer,
vs air cooled cpu.

You are technically correct, but the difference in boost clock between 60C and 80C is margin-of-error.

Quick sanity check on my GPU in a static scene with me choosing hotspot temps via fan control (Unigine Superposition, 4K, Extreme)

55C = 2745MHz
60C = 2730MHz
65C = 2730MHz
70C = 2730MHz
75C = 2730MHz
80C = 2715MHz
82C = 2715MHz
85C = 2700MHz
90C = 2645MHz

Basically nothing happens at normal temperature ranges, it's 30MHz between god-tier graphics cooling and barely-adequate cooling that's bumping up against the 83C throttle limit, more realistically it's 15MHz which is half of one percent.


This isn't remotely suprising because GPU-Z always gives the same reason for why it's not boosting any higher:

Nvidia have their partners locked down so tight it's ridiculous. By the time you've paid 25% extra for the special edition factory overclocked version where they're allowed to increase the power limits, you could have just afforded a much faster GPU from a tier or two higher in the first place.

 

[rusty caterpillar]

@Chrispy_
while that works for cpu/case fans, its not useful for gpus.

last time i checked, Nv starts throttling boost clocks at 43C for every 6C increase, so the cooler it runs (under game load), the better.
one reason you wont find any aircooled gpu that boosts higher/longer than any LC version of it, unless maybe its running outside the case in a climate controlled room.

part of the reason in recommend ppl to get an AIO, so the rad dumps the cpu heat outside the case, lowering in-case temps, thus making the gpu boost higher and longer,
vs air cooled cpu.

Not in all cases. Depends how ppl build their air coolers around their PC.
Haven't seen my card throttling down or being boosted.
I changed the leaky EVGA pads with thermal putty recently.
I run my PC usually with AIO 280 mm but testing with Mugen 6 which is not fantastic.

Look at this test, Superposition 8K optimized 20 minutes Game> Cinematic loop.
Clock stays on GPU around 1870mhz during the test.
The power however behaving strange, maybe is the Studio Driver 560 I'm using now but, power plan is Ultimate Perf.

 

[materia]

Would you have a moment to check what revision the fans you tested are? Should be somewhere on the box.

Mine have the airflow and rotor arrows on the side of the fan (rev. 4). Seems strange for them to remove it. But at the same time it would be strange if they sent you old stock.

If you ever test the P12 max could you check the stability of rpm on the lower end please? Some reviewers like hw cooling found that the motor revs up and down below ~800 rpm as it's struggling and the pwm pst outperforms it until quite high speeds. Either it's the luck of the draw or way too many reviewers have bad methodologies.

 

[rusty caterpillar]

Would you have a moment to check what revision the fans you tested are? Should be somewhere on the box.

Mine have the airflow and rotor arrows on the side of the fan (rev. 4). Seems strange for them to remove it. But at the same time it would be strange if they sent you old stock.

If you ever test the P12 max could you check the stability of rpm on the lower end please? Some reviewers like hw cooling found that the motor revs up and down below ~800 rpm as it's struggling and the pwm pst outperforms it until quite high speeds. Either it's the luck of the draw or way too many reviewers have bad methodologies.

View attachment 375038

ATM I have a Mugen 6 as CPU cooler. I still have Arctic II 280 and I will test it. I do not have P12 as obviously I own a 280 AIO which has the P14.
The oscillations in RPM are normal no matter of revisions as long as your BIOS is handling the CPU monitoring and cooling, will rev up the RPM as your CPU will increase in temps and than go down again in RPM. I do have the arrows on my fans, was just hidden as fans was mounted in that way, but still is and ever was in push, except, in the test with Rad on vertical mounting.
I guess the pump got sensors for water temp inside the loop and limit of the water temp is established by the manufacturer so that is prolly another cause for fan RPM oscillations.
However, yes, the oscillations of RPM can be annoying in terms of noise.
What I have done: I let my BIOS profile for CPU on Standard not Silent downloaded the fancontrol software and set the RPM to 1000 RPM or 1100 for the 2x P14 of the Arctic II. In that way the noise is kind of constant as a certain level and not up and down all the time.
What I know atm form my previous tests: on silent profile in BIOS with no fancontrol software the oscillations for the 2x P14 can be anything in range of 700-1300 sometimes going to 1500 RPM if I benchmark.
https://getfancontrol.com/ download calibrate set a fixed RPM for your AIO and enjoy. Note that P14 needs less RPM than P12.
Hope it helps

 

[VSG]

Would you have a moment to check what revision the fans you tested are? Should be somewhere on the box.

Mine have the airflow and rotor arrows on the side of the fan (rev. 4). Seems strange for them to remove it. But at the same time it would be strange if they sent you old stock.

If you ever test the P12 max could you check the stability of rpm on the lower end please? Some reviewers like hw cooling found that the motor revs up and down below ~800 rpm as it's struggling and the pwm pst outperforms it until quite high speeds. Either it's the luck of the draw or way too many reviewers have bad methodologies.

View attachment 375038

Thanks for pointing it out, I just noticed those arrows. It wasn't immediately obvious under the studio lights I use, I've updated the review accordingly.