What is the best placement for your heatsink fan? - Noctua NH-U14s

Hey everyone! This video was inspired by a commenter asking about if it would be possible to move the fan on the other side of the heatsink would result in any issues. The results were quite surprising (for me at least)!
I'd love to hear your feedback in the comments below! Please drop a like/subscribe if you enjoy the content that I create on my channel. It means a lot!
Heatsink: noctua.at/en/nh-u14s

[@PatrickRennie]

If you enjoyed this video, please consider liking and subscribing to the channel for more tech-related content!

I bought the NH-U14s yesterday I'm waiting for it to arrive. Thanks for the info. Greetings from Uruguay!

[@PatrickRennie]

Nice! Enjoy your new cooler 🙂

[@ChrisKadaver]

You need to set a fixed value for the fan speed. Otherwise you will get about the same result for every test since the only difference will be fan speed instead of showing the temperature difference that would've been reviled if the fan speed was fixed.

[@ChrisKadaver]

@@heartman1486 You do that in bios/UEFI or there are software to do it if you prefer that.

[@jasonking1284]

Fan placement will not affect setting fixed fan speeds or not.

[@BookFurnace]

Man, you either painfully misunderstood the comment or painfully misformulated your thought. He's saying that the experiment does not make sense since even if there is a difference in cooling efficiency based on fan positioning, we won't see it because fan speeds will adjust to compensate.

[@itsweenuqx]

I’m really glad to see this works with the fan reverse. I had to do this unfortunately. I got my hands on some good ram for a cheap price, but they are taller and don’t allow space for my CPU fan. Love to see it’s minimal with temperature change! ❤

[@superganjahleaf619]

Three yrs later and still relevant. Thnx for the vid

[@brianandrews5084]

Thanks for answering this question!

[@45eno]

5:24 actual reason it was louder in pull config and push/pull config is the pull fan placement to the heatsink. That heatsink was not designed to have the pull fan that close to it. The intake side of the fan is so close to the heatsink that it makes turbulence. Some of Noctua's coolers and many other brands need a slight gap on the pull fan to avoid this added noise. Noctua even includes thicker corner fan pads in the box for the U14S and likely many other coolers I don't own. Always pay attention when you install any fan with the intake side close to something. If you hear a new noise, don't just slow your fan down, find a way to put a gap between the fan and the surface it is flush against. Crazy this is not talked about more in tech channels, pull fans are always vulnerable to weird additional noises from the fan blades simply being too close to an object.

[@gustl1177]

absolutley right. have the same issue with my cooler master. push/pull makes noise, till i played with the gap

[@toomanylies7716]

Awesome test! Glad I don't have to buy another fan for my build 😁

[@weiposkha7428]

Yeah me too ty for this vid

[@nolanlemos]

Thanks for the informative video!

[@michaelpoczynek]

Thank you. Very helpful and interesting. 😀

[@DrBreezeAir]

If you're experiencing RAM clearance issues, just mount the fan slightly higher on the radiator. P. S. Kudos for the Studio Ghibli background.

[@45eno]

Finally someone that already posted the obvious. As long as your case can handle the slight fan height increase then you can easily make even the dual tower D15 work with stock front fan AND 4 sticks of ram.

[@jasonking1284]

@@45eno ONLY if raising the fan clears the case cover. Not a trivial point but a make or break point.

[@45eno]

@@jasonking1284 Did you even bother to read what I wrote? I already said as long as the case can handle the slight increase in height. Did you even bother to read?

[@jasonking1284]

@@45eno Yea. I did read. That's why I wrote the message. You need to remove the word "slight" because it trivialises the situation whereas the reality is, even a SLIGHTly raised cooler is going to be a huge problem if it doesn't fit. Either it clears the case cover or it doesn't. "Slight" does not come into it.

[@armandpeanutspinou3430]

Yeah not possible with mine unfortunately

[@TraveliandoCo]

What about the RPM? It is okay to have the same rpm on both? Or would be better to change it?

[@medoug25]

Great vid, was worried specifically about the loss of cooling because one fan would block my case cover, really needed to know if only running one fan attached to the heat sink would impact me at all. Glad to know there’s not much difference

[@MrEricmopar]

The results are probably because the dual fan setup was done last and the system was fully heat soaked from running longer. Ergo the system slowed some. I have a U14S on my Ryzen 7 5800X build and the Cinabench R23 points drop, compared with doing a test when the system is just started and cool, to when it's been running for hours and heat soaked. (all background programs closed, except security) Other's that have done these dual fan tests, on single tower coolers say the same thing, more noise, no increase in cooling. (In fact in a lot of test cases, the dual fans increased temps) the reason for that is, that not all fans are created equal. The exact same model of fan, can easily be plus or minus 100rpm from the specs, or even worse with cheaper fans. Even Noctua's newish and vaunted NH-U12A runs the same with just the front fan and no rear fan.... (but with less noise)Hmmm. I have enough experience with Noctua now, to know they make excellent fans, but like every other company, there is a lot of hype around the coolers themselves. For instance, the solder between the heat pipes and base should be visible but isn't. On the NH-D15S the cooling is so degraded by the single fan, you are better off getting a single tower solution. Either that or plan on the increased expense of adding a 120mm fan to the front tower on the D15S and running the middle fan in a push configuration on the second tower.

[@18bagabooo]

that's a good video. I just had to do the back-fan placement single due to ram clearance issues and I was wondering if is going to affect the performance. I was even thinking to add a smaller 120mm fan to the front. but seeing my own temps (vs old cooler) and this video, I think now is no point to worry about.

[@PatrickRennie]

Glad you got something out of the video!

[@AKaptijn94]

I'm going to order the NH-U14S soon. I have the same case btw.

[@PatrickRennie]

Nice! The case has been amazing for me for the past year. It's so easy to work in!

[@TheZanzou]

Thanks for this, the fan blocks one of my DIMMs so I was wondering how much performance loss I'd get if i switched from push to pull, answered my question perfectly

[@45eno]

Just raise the fan higher. As long as your case has enough room you can still have all 4 sticks installed and use the included 150mm fan. I am doing this now on one of my PCs.

[@TheZanzou]

@@45eno eh already made the swap from push to pull but that change probably would be the same if not more cooling loss although a lot easier to do.

[@45eno]

@@TheZanzou Yeah it's a minor degree difference on push vs pull but just saying you can easily raise the fan higher on the heatsink and still make whatever RAM config work. The D15 with a push fan overhangs almost all the ram slots and my daughters PC has 4x8gb installed and easily accepts 2 fans on the dual tower D15 heatsink. Just letting people know it's as simple as raising your fan to make whatever RAM config work. As long as the case has enough width to support the fan raising up a bit past the heatsink.

[@bofasofa9399]

@@TheZanzou ?? Why did you say "eh already made the swap" as if it doesn't take literally 15 seconds to add or switch a fan's position? lmao unbelievable levels of laziness

[@TheZanzou]

@@bofasofa9399 because the configuration it's in currently is likely as effective if not more effective than the potential solution that they offered. Putting the fan so it's hanging off of the heatsink means less airflow through the fins which less cooling potential. Me saying "eh already made the swap" was me saying my problems already been solved so there isn't any need for their suggestion.

[@cristianroth8524]

I have a SilentiumPC Spartan 5 cooling an i5-12600K. Ever since I mounted a 120mm Noctua Industrial 2000, the temps dropped by 10-15 degrees. So, it's the speed of the fans that makes the difference. By mounting two identical fans running at the same speed you'll get other advantages as well: the cooler will have a slightly better "reaction speed" (it will start cooling faster) and by forcing out the same volume of air that was pushed in the heatsink, it will help regulate the system's general airflow.

[@jasonking1284]

That fan must be loud...

[@cristianroth8524]

@@jasonking1284 It is when pushing beyond 80%, but nowhere near as loud as the 3000 RPM version. But the performance is worth the extra noise, trust me. I plan on upgrading to an AK620 and I think I'll do the same (replacing the fans, that is).

[@jasonking1284]

@@cristianroth8524 Interesting. Just installed a Peerless Assassin which replaced a Hyper 212. Improved temps by -20 degrees. Very impressed. Might get a high speed fan to replace the standard fans and adjust the fan curve when I upgrade the CPU..

[@remaincalm2]

Interesting, thanks for sharing your experiment. I'm thinking of buying the same Noctua fan for my Ryzen 5900X (pre-ordered) so it's good planning info. I wonder how well the Noctua can cool this new chip? I decided not to get a windowed case because the Noctua isn't attractive without any fan LEDs to compensate for it! You have a very clear, calm voice, easy to listen to and well suited to conveying information to people. I know you didn't ask for feedback, but I figured you wouldn't mind a compliment! Good luck with your channel.

[@PatrickRennie]

Hey thanks Gary! I really appreciate the kind words :) 5900X eh? That's going to be a pretty powerful CPU! I can't speak for how well the Noctua would cool that sort of a chip, but I can tell you in my day-to-day use of the Noctua + my 2700X, temps will rarely get above 65 degrees (and I usually see that temp when I'm rendering a video for KZbin). The stress tests will take it closer to 70, but overall I'm really happy with the cooler. It's definitely not the 'prettiest', but for the performance + lack of noise in the system overall, I'm pretty smitten.

[@remaincalm2]

@@PatrickRennie It sounds like there's plenty of head room to keep the 5900X far below the max temp of 95 deg! Based on the temp you get when doing encoding I'll gamble the Noctua will keep the 5900X below 75 doing the same thing. One day, when stock becomes available to ordinary people, I'll find out. ;-)

[@osmantopcu1660]

Hey Gary what is your current temps with u14. I have same specs, 5900x and u 14 and temp goes up to 90(mostly 82 C). I am thinking buying second fan to decrease the temp but I dont know. Maybe u14 is not good for 5900x.

[@remaincalm2]

@@osmantopcu1660 That's disapointing to hear you're getting temps up to 90. Is 82 at idle / light use? Is there enough ventilation in your case and does the hot exhaust from the u14 have a direct path out of the case? Consider using a case fan to suck that heat out. Sadly I haven't got my 5900X yet, no stock. AMD supply problems. :-( Everyone wants one!

[@draco3984]

@@remaincalm2 so how was the degrees?

[@jhendra83]

you can also use a thin fan120x15 fan. I did that with my 212 black edition.

[@fred_2021]

I've tried both single fan configs. In my system, there's about 1c temp difference. Both are inaudible. Phanteks P600s, 3 x Arctic Bionix P140 front intake. 2 x Arctic Bionix P140 top exhaust, 1 x Phanteks stock 140mm rear exhaust. i7-10700K 5GHz. 3080 OC Different cases and contents might give different results, but I suspect most users will see hardly any loss of cooling efficiency, while having ~ 2cm clearance of their 4 x RAM RGB.

[@jayc0au]

I installed at the back and was questioning the temps. I’ll switch it to the front later since the MB has plenty of room. Thanks for the info.

[@PatrickRennie]

No problem! What are your temps like as it's configured right now?

[@jayc0au]

@@PatrickRennie I literally finished my 5600x build yesterday. I’ll measure the temps before I switch the fans to the front and report back.

[@jayc0au]

@@PatrickRennie After 10 mins OCCT with Back: max 76.1C avg 70.3C Front: max 71.8C avg 66.2C Please bare in mind when I made the switch from back to front, I had put less thermal paste on the CPU.

[@halosxsjk]

@@jayc0au can I ask that when using the fan in the back, was the fan pusing air into the heatsink or pulling air away?

[@jayc0au]

@@halosxsjk I have the noctua U9. By default, the fans pushes air into the heat sink. All my test are done on closed case, stock clocks, stock MB settings with ambient temps of 28c.

[@suatsekil]

great work thanks

[@PatrickRennie]

No problem, thanks for watching!

[@MrLincoln87]

I got a U12A cooler which comes with 2x120 fans. I had already factored into my build that I was going to remove the rear fan and use it as my rear case fan (meant 1 less fan to purchase too) This rear case fan is bridged/split from the cpu fan header too =) Liked the testing, wish I had the time and patience to actually test it all out.

[@PatrickRennie]

It's interesting, before getting into the hobby I would've assumed that tactiles were the most popular. I guess people value smoooooooooth over a nice bump.

[@bofasofa9399]

What is with you people? Why would you buy the U12A just to remove a fan and use it as a case fan. Disgusting.

[@MrLincoln87]

@@bofasofa9399 I forget where the testing was, but there was a minimal loss to temps by taking it off the back. Not to mention, I don't have the room to have a rear case fan and a rear cpu fan sooo.

[@OfficiallyOrca]

In my experience, the pull method (after the radiator) will make the fans work harder which makes them louder. Having the fans set up to keep stall (neutral air) from building up is always a plus. No matter, if you prefer AIO/Air, proper air circulation, if you prefer AIO/Air, proper air circulation will save you a lot of money.

[@kuntilanakhybrid4043]

I directed both CPU fans to the center. and directs the rear fan of the case inward. I installed 2 fans above the casing for exhaust output and 2 fans above the PSU leading to the GPU. I get cooler temperatures. and I installed 3 fans pointing inward at the front of the casing. so the heat dissipation path is only at the top of the casing

[@nal_er]

Makes sense. The back flan is actually slowing the air flow out of the front fan.

[@45eno]

No, the problem is the rear fan doesn't have a proper gap. In the Noctua U14S box they include thicker rubber corner pieces for their fans to sit back a couple MM from the heatsink stopping the weird noise. Suction / Intake sides of fans always have a good chance of audible noise due to the fan blades being too close to the object. All he needed to do was use those spacers and the noise issue would have been fixed.

[@robertkubrick3738]

Twin tower cooler both facing in towards the center with a side door fan as exhaust.

[@xq0404]

Thank you as I was considering buying a second fan for my NH-U12S TR4-SP3. I have now decided not to do so. Adding another fan may only marginally reduce temperatures, but increase noise to quite some degree.

[@PatrickRennie]

Yeah it doesn't seem worthwhile, at least with this particular cooler. Glad you found it helpful!

[@PavloSmal]

In my case with TR 3970x and same cooler adding second nf-a15 fan dropped temperature significantly.

[@PavloSmal]

Sorry, I have U14s-tr4 actually

[@45eno]

You won’t get this noise penalty if people used the supplied thicker bumpers for the pull fan. I noticed this noise as well on my setup and then realized Noctua already provided a solution that comes in the box of the U14S.

[@ultralaggerREV1]

Hello, I am deciding between different Noctua heatsink cpu coolers. I will have my PC moved to the NZXT H6 flow, and I want to optimize air flow to Positive pressure. I will have the stock 3 fans for intake, but I want to make sure I get most of the hot air out to get cooler air as quick as possible while reducing dust collection

[@donkee]

Checking the CPU fan RPM it seems as if it's at its lowest when having dual, but performance wise and noise wise you probably get a better result with the front only.

[@Kaffeestrudel]

Thx for that test, I´m stunned that is nearly no dif. betwen 1 front and dual fans, only you pay less because no +1 fan to buy. After that I would be intrested how the Noctua NH-U12A is vs this Noctua in cooling.

[@jjglaser]

I am about to put my u14s on an am5 cpu. i'm sure it will do just fine. Good to see theres no reason to add an extra fan.

[@melodychest9020]

Nice one. I think your rear PC case fan is close enough to keep pulling the hot air coming out of the heat sink that the back fan is not really doing anything much extra? Yes, these Noctuas are built good and one fan should do it as they are pretty large enough anyways.

[@anythingbenz4005]

When having two fans, if the back one is spinning a bit slower than the front it could block airflow possibly or create a hot air pocket?

[@45eno]

Use a 2 way splitter and run both matching fans to the same fan header. Noise on the rear suction fan is from improper spacing of the suction fan to an object.

[@vdfritzz]

with cheaper fans and heat sinks, having 2 fans improves temperature a lot in my setup it was nearly 10c noctua quality is insane

[@AlehandrosArhangelos]

If you put your 1 fan to the end it needs to suck air obviously to throw it out of the case. But with this cooler you always get 2x fans. You do get better performance, I have both and you get -2C ~ -5C with 2 fans vs 1 fan.

[@PatrickRennie]

You got 2 fans with this cooler? The NH-U14s only comes with the single 140mm unless they've changed their product. I had to buy a 2nd 140mm for this test

[@davapod]

I did the same test and found the best config was with the fan at the front. I also tried the dual fan set up and temps dropped a mere 1-2C. However the dual fan set-up had a distinct increase in audible noise. To the point I thought there was a fault with my second Noctua fan, swapped and tested, was fine. Possibly an interference issue. My experience says a second fan adds near to nothing. Hope this helps.

[@RossBoss90]

A couple C is a a couple 100mhz on a OC

[@davapod]

@@RossBoss90 For sure, in my case it was about lowest temps whilst being quiet with the standard clock. I do see your point of view though.👍

[@45eno]

You only needed to use the supplied thicker fan bumpers that came in the box to get rid of this noise. Same thing happened to me when I added a pull fan on my D15. Those thicker bumpers only come in the U14S not D15 as the U14S was meant to have a pull fan added if desired and Noctua knew there is turbulence when the pull fan is used without a slight gap on it. If you look at your fan you will see the fan blades are closer to the edge on the non motor side.

[@davapod]

@@45eno Thanks for the input. I can't find the original box, got lost in a house move. I have a couple of chromax fans on order, any idea if these use the thicker bumpers? 🤔

[@miadu2466]

should both fans be facing towards the exhaust fan of the case or should they be alternate?

[@TransformX]

You should redo your test. The fan from the back needs the thicker pads against the heath sink. Check their mounting manual. It could help for the temps, because of the space needed for the rear fan.

[@PatrickRennie]

Should clarify; thicker pads were used after I double checked. The video shows the old pads on as that was the only footage I had, but the results are with the thicker pad adding the additional space between the fan and the heatsink.

[@drdank9013]

This is an important tip, using the fan on the back as a temporary fix (considering sand paper modifications so the ram will fit). Additional question: in my pc, the fan will fit in the front of the heatsink but it very slightly bends the ram in order to do so (nothing beyond elastic limit) obviously, I want to avoid this but how much of a problem would it actually be if I let it?

[@robertkubrick3738]

@@drdank9013 As long as the fan doesn't vibrate? Nothing. If it did vibrate, I think it could possibly damage the ram slot over time.

[@45eno]

It's crazy to actually see someone mention these thicker pads. I see so many people mention louder noise when using push/pull or simply pull configs and 99% of people do not understand that pull fans have their fan blades too close to an object causing the noise. The advice I have seen is to simply lower your fan RPM to counter it. The correct solution for Noctua at least is the thicker pads on the fan to gap the fan and heatsink just a bit to eliminate that pull fan noise. This is a common fan noise issue anytime running a suction/pull fan too close to a case or heatsink.

[@KamenMinkov]

It would've been useful to either show graphs of the speed of the CPU fan[s] or to fix them to a constant 100% - otherwise, if they're always targeting a certain temperature, they will differ in their rate. I also have the U14S and I've found that, weirdly enough, at full CPU load, there's only like 1-2 degrees difference between 80 and 100% fan rate.

[@PatrickRennie]

Hey that's great feedback! Thanks for your response. I'll be sure to take that into consideration when I'm doing more temperature/fan based tests in the future.

[@L3nny666]

yeah i never let it run over 75%, so around 1000-1100rpm.

[@AzzulyX312]

Hello so for the first time in my life im having a heatsink fan and lots of time the heatsink fan will get loud its on the front side i have 2 frontside intakes and 1 backside exhaust the heatsink fan keeps being loud

[@Thats_what_she_said__89]

I got the ryzen 7 75700g would this cooler work good with that CPU?

[@MagicPlants]

nice

[@PatrickRennie]

Nice

[@ADP3000]

I remember reading somewhere that in dual fan configuration, the back fan must be spinning faster than the front fan. Is this how you configured your test?

[@PatrickRennie]

Hmm that's a very good question. The way I had my test set up was that both the front and the rear fans were connected to the same fan header via a splitter. This would have meant that in my test the fans would have been spinning at the same RPM. I think it would be interesting to try this experiment again but have the two fans on separate headers to see if that makes any noticeable improvements. Too bad I sold the second fan haha, I didn't think I would need it anymore :P

[@italoop7850]

@@PatrickRennie Actually NOCTUA recommends using a faster fan in front and a slower one in the rear of this cooler. The fan that comes with the cooler has a max spinning rate of 1500 rpm. The regular 14mm fan that Noctua sells for the rear position tops at 1200 rpm so even when connected to the same plug, using the included power splitter, there will always be a 20% offset between the two fans speeds. That's how I use mine with a Ryzen 5 3600. Temps are better with 2 fans. I'm getting a max of 62.5° on a 40 minutes pass of Prime95 stress test and CPU fans run at max 1100 front/about 900 rear as that's the sweet spot for performance/noise. My case is a Fractal Design Meshify S2 with 2 14mm fans in the front, 1 14mm fan at the bottom as intake and one 14mm in the back as exhaust, so 6 14mm Noctua fans in all. With a single fan on the front of the CPU temperatures are higher of about 2 or 3 degrees than with 2 fans running at offset speeds as Noctua suggests. Not a huge difference but this keeps fans at lower speeds and noise is ok.

[@45eno]

@@italoop7850 You only needed to gap the rear pull fan to eliminate that weird noise on the pull fan. Noctua makes thicker fan corner bumpers for this very reason. I run all 3 of my fans on. 3 way 4 pin cable and they are spinning approximately the same rpm. Super quiet. The noise comes from spinning fan blades too close to the heat sink.

[@mot.studios]

Nice

[@PatrickRennie]

Nice

[@zealoustherealous]

vibration my guess is the electricity flowing through the cpu was effected by the vibration of the fan which in a way can be considered energy. or the electrical field generated by operating 2 fans created interference.

[@robertkubrick3738]

You might check if the fan in pull configuration was running faster.

[@ralre666]

Is the heatsink in contact with the gpu?

[@EliteRock]

Currently running Noctua NH-U14S. Both my last two builds, I've had the CPU fan at the rear of the cooler blowing forward. The top-rear case fan is reversed and pulls air in through a fine mesh filter on the outside (magnetic, can be removed and cleaned in seconds). A centrally mounted 14cm top fan is perfectly positioned to extract the current of warm air coming off the front of the cooler. BTW, The NH-U14S didn't actually improve temps from the Thermalright Ultra eXtreme 120 it replaced, but it is quieter. That rear case fan is a 12cm BeQuiet Silent Wings which has a circular duct shape onto which which I've mounted an 'eggbox' grille made of black card to effectively make it a 'vaneaxial' fan, the vanes straighten and direct the flow onto the NH-U14S's fan (they also prevent any weird harmonics which can arise when fans are effectively 'stacked' this way). So, the CPU is cooled directly by air pulled from outside the case. Works extremely well, keeps all cores in a CPU pulling 125+ watts (c. 220w at the wall) at 70° or less, most in the mid-60's (running Prime95 AVX, small FFT's). With real world tasks like x264/x265 encodes, the NH-U14S's fan and the BeQuiet case fan (slaved to it) never need to spin up much over 800 RPM to keep all cores at c. 60°. ETA >> basically this arrangement has the advantage of not using air already warmed up by the graphics card et al to cool the CPU - I've run MSI Combuster (overclocked GTX950) simultaneously with Prime95 (resulting in well over 300w at the wall) and it makes no difference to CPU temps (the air coming out of the top exhaust fan gets REALLY warm). There are also two front-mounted and one centre-bottom 12cm intake fans (all intake fans are filtered and the case overall is 'positive pressure').

[@maxheadroom6833]

A good video! Which would preform better, a U14 (2 fans) or a D15 (2 fans)?

[@PatrickRennie]

Hey thank you! I think that the D15 will outperform the U14, because you also have to consider that the D15 has 2 stacks of heatsink fins, when the U14 only has one. Keep in mind though that the D15 is a huge chonker, and you'll need to make sure that your case has the space for something that large. I opted for the U14 as it was cheaper, and I don't really dabble with overclocking and what not. Hope that helps!

[@maxheadroom6833]

@@PatrickRennie Thank you for the reply! I opted for the NHD15, as sadly Noctua dont make and have no plans to make the U14 in black

[@PatrickRennie]

Nice! I hope it works well for ya 🙂

[@xxMIK1989xx]

@@PatrickRennie I have a question too, which one would be better? U14S with 2 fans or U12A? Thanks.

[@Wigglythegreat2]

Due to case restrictions, I have to use the smaller Noctua U9S. With the smaller 92mm fans, do you think that I might benefit more from two fans? The cooler came with one and it greatly reduced my CPU temps already, but I am wondering if it is more likely to make a difference because of fan size.

[@45eno]

Old post I know but yes two fans can help lower a few degrees. I have a U9S in a media pc running a 5800x (hot cpu) and I decided to put the extra fan on this setup for the slight cooling boost. I also have an almost identical media pc in the bedroom but with a 5700x and a D9L and single 92mm. I could add another 92mm fan to this tiny dual tower cooler but it won’t give me much more performance and I want the most silent bedroom setup.

[@NGreedia]

I find having a fan in pull config isn't desirable as it creates an annoying acoustic. I have a Deepcool AS500 plus and at full speed, the fans are only at 1200rpm but the rear fan creates that loud noise due to it trying to pull air through the heatsink. If I run the fan at max with just the front fan, I don't really hear it as much. This is why for intake fans, it's best to have a gap/space between any mesh panel and the front of the fan. If you put your hand directly in front of a spinning fan, you'll hear the difference as well.

[@ortizma13]

The voltage used for the second fan probably affected your render performance slightly

[@nickgarcia610]

Does it matter which direction the airflow goes over the heat sink.. like should the air move through the heat sink or push the hot air away from it? I can’t figure out what to do

[@vLarsn]

Ideally you want your fan to push air through the heatsink. You always want airflow to go in the same direction. So typically you'd have your front and bottom(if you have those) case fans blowing air into your case and your back and top case fans blowing air out of your case. Keeping the airflow in mind you have 3 configurations(as shown in the video): 1: Pushing air through the heatsink with one fan on the right side. (right=front side, left=back side) 2: Pulling air through the heatsink with one fan on the left side. 3. Two fans. Right side fan pushing and left side pulling air. Ranking the different configurations from best to worst it would be 3-1-2. Option 1 and 3 utilizes the cool air from the front and bottom case fans the best. I probably overexplained this a lot, but I hoped it helped.

[@flaminggold7593]

Is a 2 degree decrease that beneficial for dual fans?

[@PatrickRennie]

In the grand scheme of things I don't think it's a huge difference, no. I think if it were 4-5 that would be more significant. That being said, this is only one test. I'd be curious to see if there's any changes with the U12s (the smaller version of this) for example.

[@Ballacha]

Push & pull is only good on air flow restricted setups, for example through a radiator, or if the intake fan has dust filter on it. Otherwise it’s a waste of money and energy for negligible to straight up negative performance gain.

[@logorex]

Negative performance? How so?

[@Ballacha]

@@logorex go watch jayz2cent's tests

[@ihaveasmallpenis4983]

Will This let me Download More ram? Where can I download More ram?

[@PatrickRennie]

You can download several megabyte of RAM with this cooler yup 100%

[@ngocsonphamba]

Sir, NH-U14S vs NH-D15 who is the best? Im use ryzen 7 2700x Sorry im bad english

[@PatrickRennie]

I think it'll depend on your case size - the D15 is quite a bit thicker, but will do a better job cooling down the CPU. I'm still using the U14s in my build and I find it to be fantastic. Currently just running with a single front-fan and I'm not having any issues!

[@ngocsonphamba]

@@PatrickRennie thanks you sir !

[@XMoldyMe]

If u have big Noctua cooler. Go to 2 push fans and 1 on top for heat coming out.

[@Sasaking999]

Than can’t be good? You’d have the airflow cancelling each other pretty bad

[@allroundgamer3541]

so i found a gaming pc that was thrown away like on the street with a strix 1070 asus maximus hero viii 100wevga fully modular psu 4x8gb corsair lpx vengance ddr4 andi it has the same cooler how well will it perform since i dont have this blender test

[@PatrickRennie]

Lol that's rather good fortune if true. You can download the benchmark tool directly from Blender's site, it's free: opendata.blender.org/

[@allroundgamer3541]

@@PatrickRennie yup true as can be as I'm 15 and can't afford that type of pc lol

[@matthiasmcnulty407]

This is key info.

[@PatrickRennie]

Thanks! It was a fun little experiment

[@RamaOshiLo]

i have this one too, and i placed the fan at front. there is 3 fan : 1 big fan in front case for intake, 2 small fan in back case for exhaust.

[@buggerlugz6753]

Could you do one with both fans blowing into the heat-sink, essentially forcing twice as much hot air out the sides of the sink? If your external rear fan blows into the case too, it'll exhaust the hotter air far better out the top of the case.

[@Knottypine79]

I don't think that would be very efficient. The idea is you want a steady flow of fresh air passing through your case and over your components. Facing all fans inwards except for the top wouldn't really be an ideal set up and would more or less create a windstorm inside your case. I believe the ideal set up would be having intakes on bottom and front, exhaust on rear and top. Facing two fans on a heatsink both exhausting inwards would not provide sufficient cooling over the fins.

[@CaptainCiph3r]

I grabbed this fan for an I7 9700K on recommendation from a friend. Hope I didn't screw that up

[@PatrickRennie]

I don't think so. They're excellent fans regardless, and based on some feedback in the comments I'm getting the two fans should be on different headers. The rear fan should be spinning less than the front to create less turbulence within the fin stack by the sounds of it. I might revisit this test in the future because it sounds like there are a couple things I overlooked that I want to look into. And even if the performance is marginal, you still have a great fan that could be a case fan somewhere else in the build!

[@darkfoxjj]

all fans have to be synchronized otherwise they might block each others airflow slightly

[@helthuismartin]

In the middle

[@mahmutdikcizgi9773]

Whats the point of noting max temp, you should compare average since the purpose of the fans is to dissipate the condesated heat form the heatsink. The max temp is mainly related to your heatsink and its thermal conductivity. And please use prime95 small ffts for heat test, then you will notice the diffs.

[@mahmutdikcizgi9773]

And set you cpu fans to constant pwm/speed so you can compare oranges with oranges

[@randybjake]

I would remove the back case fan, now make sure that both fans are pushing to the back of the case where you removed the fan . Now re run your test.

[@45eno]

Thanks for the well made video. But I have answers to your conclusions. Except the render difference. That would only be if temps were higher then less boost clock on cpu. 1.) The reason it was louder anytime you had the rear fan installed is due to the fan blades being to close to the heat sink. The fans motor mount arms place the fan blades enough distance away from the heat sink when in a push configuration but once the front of the fan is placed up close to the heat sink the audible noise occurs due to turbulence. Use the included thick bumpers that come in the box for the U14S which I own as well. I just made a quick video showing this. 2.) Even with the D15 the dual tower version you can use the fan and 4 ram sticks. Just clip the fan slightly higher. Have this setup running in a Corsair 4000D airflow case with Black D15 and two 150mm fans. RAM is Gskill FlareX with taller than low profile heat sinks. 3.)It’s actually a 150mm fan in a 140mm package. I run 3 of them on a D15 with a 5900x.

[@ciliper8492]

Bro how is better ?

[@Stuntzii1]

i actually have mine sideways which saved my computer heres my story the noctua and the asus mother board i cant believe had no damage I can say that the mounts are super durable. I can't believe my computer is not dead. So.. i was moving out of my apt and taking my bed apart which is older and all heavy hardwood. i keep the side off my case btw. i was picking up the sideboard of the bed frame and the headboard which was leaning against the wall fell over onto the sideboard that i was holding which knocked it out of my hand and hit directly onto the noctua cooler in my case with no side. i was like aahhhh! im glad to say its been over a year and its running fine. it did bend the heat pipes a little but no cpu or motherboard damage. asus mb btw. im very impressed with the mb and the cooler

[@Wigglythegreat2]

Built Noctua tough lol. 😎👍

[@MagicPlants]

It's because the fans were fighting eachother

[@PatrickRennie]

😂😂

[@highcel]

@@PatrickRennie He was not joking :) back fan should run at lower speed.

[@jamesm568]

Doesn't really matter as it comes down to the characteristics of your case.

[@ZEROZERO-sv1cs]

You tested front first and as you proceeded you probably didn't let the pc cool down enough. But no major difference there

[@tequilaghoul]

what's the height of the cooler?

[@PatrickRennie]

Height is 165mm according to the Noctua website. In the Fractal Meshify C (my case) you'd have maybe 10mm of clearance between the heatsink and the tempered glass.

[@tequilaghoul]

@@PatrickRennie thanks for the fast reply. Was looking to get this for o11 XL but it would be soooooo close. XL has 167mm of clearance. I’d like the ease of building in that case but I wouldn’t mind going with phanteks 719. Sure it’s bigger, but has clearance for cooler and I can add more fans for more air.

[@PatrickRennie]

Wow that Phanteks 719 is a super unique looking case! If clearance is an issue you and you're still looking at a Noctua, you could also get the U12S instead of the 14: noctua.at/en/nh-u12s/specification. Clearance on that one is 158mm. I installed the U12 in my girlfriend's PC (Phanteks P400A case), and it also works super well.

[@tequilaghoul]

@@PatrickRennie I found this cooler RIGHT after I commented! This cooler totally gave me the ok to buy the Lian Li o11 Dynamic XL! even if I woudnt have found it you would have pointed me in the right direction! You are epic! Thanks!

[@luisalvarezfelmer4284]

@@PatrickRennie HI there hv a question would the heatsink cromax covers fit... was wondering cause its states thtat will rise up 7mm...any ideas.... meshify c states 172 mm max cooler height if am not wrong...have yo tried???

[@heighRick]

It's simple ...move the fan up the front side??

[@elitexcomputing8802]

Pop that rear exhaust fan off of there, will drop your noise so much!

[@PatrickRennie]

Interesting, I hadn't even considered that. I would assume that would increase the latent temperature in the case by a few degrees though right?

[@elitexcomputing8802]

@@PatrickRennie it should not, if you have positive case pressure the intake should push the air out.

[@joeybtheman1906]

mine is blocking 2 ram slots how in the fuck do i fix this?

[@sertac6588]

As my seing in example 2. and 3. your fan positions all wrong. If you are placing fan on back plate you need to turn case exhaust fan positon opposite. For 3. example you need to place 2. fan as exhaust and case fan position exhaust again.

[@MrWaheedulHaque]

i dropped 20 degrees after installing a second fan but then again my processor is extremely powerful so its expected haha

[@Ji-oo1tt]

Front. Back.. what's the difference when we're talking about symmetrical heat sinks? My question is would double pull config work? Edit: maybe there is a baseline/norm that I'm not used to because I'm approaching my first pc build. Now judging by what you said, 'front' configuration insinuates pushing air through, which again, I think is a weird way to look at it; because looking at the top of the cooler, the heat sinks are symmetrical meaning pushing air through from either end would have similar results depending on where that air is being pushed. So im wondering would there be any benefit to doing a double pull config? And now ill finish the video hahahaha Edit # 2: again, when you refer to 'front/back' you're referring to air being pushed out of the front of the fan, meaning 'front' configuration puts the sink on the side of the fan pushing air. So again I wonder what dual 'back' configuration would yield. I belive you only did front/back configuration, as you call it, meaning both fans pushing air in the same direction, one pushing through heat sink and the other pulling. So try dual 'back' configuration as you so confusingly, I mean, aptly put it 🤣😉

[@PatrickRennie]

Hey Benjamin, haha thanks for your play-by-play! I enjoyed reading it :P Yeah all tests were in a pull/pull configuration, drawing air from the front of the case and pulling it through the heatsink fans. I'd be curious to give your theory a try at some point in the future though! It's harder for me to make videos about this sort of stuff as it's my main editing rig (and only PC), but I'm definitely going to do some preliminary testing to see if it makes a difference.

[@Ji-oo1tt]

@@PatrickRennie I didn't mean thru case. I meant pulling air from front and back of heat sink

[@Ji-oo1tt]

So you're drawing air in thru the narrow profiles of the heat sink. Constricted air seems to cool better than a larger volume, no? Like breathing vs blowing on ur hand? 😶 am I retarded?

[@PatrickRennie]

Oh I see now. I'm not an expert but I don't think that would yield the results you're thinking it would because you're just having the same air circulate through the heatsink fins that way. Might be a fun little experiment to confirm though

[@dmitrir5529]

Your radiator is very thin. And fans is big. Thats why you don’t have difference. With a radiator 2 times bigger you will see it.

[@PatrickRennie]

I think you're right. The coolers that have two fin-stacks would be more efficient in that regard. I'd be very curious to try one of those in the future and compare to this test. Thanks for your feedback :)

[@HarryUnchained]

i think it's obvious of what type of heatsink he was using in this test. The NH-U14S. So not sure why you're asking him to use a larger heatsink. That would be a different cooler no?

[@dannynguyen2050]

Cpu fan should be installed in the front for maximum performance. SMH

[@marcinsmaga3441]

It is very interesting how pretty much everyone is doing air cooling wrong. Because of design of modern cases, with power supply at the bottom) and video cards blowing air inside, one will probably achieve much better cooling results doing the complete opposite of what most people is doing. If you setup you intake from the top and back and exhaust at the front it will be much better at cooling overall.

[@discipulodericagames]

Hot air goes up, dumb comment.

[@marcinsmaga3441]

@@discipulodericagames Very knowledgeable response. So what that hot air raises.... are you relying on passive cooling or fans to move it?

[@gducioevdgxyeuvddd3828]

You do realize you have 140mm fan in the heat sink and a 120mm fan for exhaust. The 140mm fan is bottle capped by that 120mm fan

[@PatrickRennie]

You don't think that I realized that there are two difference sizes of fan in my computer? The case I have only offers space for a 120mm exhaust fan. In my experience this is the case with most cases unless you go larger than ATX. Also from the research I've done - having a single 120mm exhaust is generally all you'll really need before you hit diminishing returns.

[@gducioevdgxyeuvddd3828]

@@PatrickRennie false. The intake cpu and exhaust fans all need to be of the same dimensions or your testing will be flawed. I know the difference having 3 140mm fans on an nh-d15 with a 120mm exhaust fan vs 140mm exhaust fan. The temp difference from what I have is about 10c cooler with all 140mm fans. I’m not even gonna debate this any further

[@PatrickRennie]

@@gducioevdgxyeuvddd3828 I'd be interested to see where you got that info - shoot me a link! I'm referencing the Linus Tech Tips vid here - kzbin.info/www/bejne/boDQnKCiqqaMd7M Wasn't looking for a debate - there's a lot of info about PC building on the internet and I'm always curious to learn more.

[@gducioevdgxyeuvddd3828]

@@PatrickRennie there’s no debate to it. My system has 3 140mm fans on my nh-d15 and a 140 mm exhaust fan. The temp dif is 100% noticeable when your case has all fans same exact dimensions and rpm. And ofc noctua fans do actually move air better than any other fan on the market. Linus tech tips….pfffff they take money from companies to “make” the results of their testing. Never rely on just KZbin. Always test for yourself. You can argue “static pressure” about dif fan size , rpm exc.(although static pressure is pure bs… more like a pc with a breathing problem IMO)

[@ace2125]

Next time, can you turn the stupid music OFF while you are talking? I COOULD NOT HEAR YOU. Do teachers play loud music in a classroom while they are instructing?

[@pryan6049]

Thanks for the test, but I will point out your mistakes. You need to show the fan rotation speed in the test; the final temperature and noise from the fan depend on this! Next, you connected the rear fan to the radiator without any gasket, which creates turbulence and causes more noise from the fan at the back! Included with the cooling kit, the Noctua comes with thick fan gaskets, specifically for installing it behind the radiator! You need to know this! By increasing the distance between the rear-mounted fan and the radiator, the noise from the fans is significantly less, and the system performance will be higher! Read the instructions for your cooling, which comes with the kit)