Front mount vs Top mount, interesting results...

[@DarkosLab]

I run my rad in the front, because it did not fit in the top.

[@JohnSmith-b6n]

Same

[@timmypowers5926]

Also same had a 240 on the top and got a 360 and to move it to the front

[@gregoryfolsom7882]

Bingo. Mobo/case combo left me maybe 3/4" between the case and the edge of the RAM. No way in hell a rad alone was fitting in there, let alone adding fans.

[@JahonCross]

get a bigger case

[@rynesherman9099]

Radiator Offset Mounts

[@BrunoRaven]

From a fluid mechanics standpoint, the main difference between push or pull fan configuration isn't just the "resistance" of the air passing through the rad before being picked up by the fan. It's also the fluid flow characteristic. Typically, the flow on a pull configuration is laminar, versus turbulent on a push config. And when we talk about heat exchanges, turbulent airflow is a requirement, as you need as much fluid molecules hitting the surfaces of the heat exchanger as possible, so they can pick up more energy, thus enhancing it's efficiency.

[@Mickelson1337]

Thanks! Makes sense!

[@mckidney1]

This is something that is not obvious to many advanced builders. This is also the reason why case airflow cannot compensate for lack of part cooling.

[@theAessaya]

There's also the sound difference. In pull configuration the fans are working against a vacuum (reduced pressure) on the radiator side, and thus are encountering less resistance, and as a result spinning faster. Also the airflow is consisting of multiple small air currents interspersed by small pockets of still air (courtesy of the radiator fins). All this cause the fan blades to generate higher pitch noise that is much more easily audible, and annoying.

[@TheManofthecross]

calm down my guy.

[@arc00ta]

@@TheManofthecross The irony of your name rejecting science is beautiful.

[@watercannonscollaboration2281]

SFF PCs: it doesn’t matter where it goes as long as it fits

[@DMFP42]

Honestly, that's how it is for pretty much anything that's not a super tower 😂 At least, if you have hard drives that aren't NVME or more than 2 SSDs

[@TacticalShazli]

that's what she said

[@ELCrisler]

SFF PC if properly built do not need an AIO as they can more easily use directed air flow. Bigger cases use air flow as a brute force proposition. SFF build allows for directed air flow to target needed areas very specifically. This means they can, with proper planning, be MORE efficient at removing heat.

[@maflarson1988]

i just took my side panel off (raws1) and slapped a Noctua u12-s on to keep cool

[@humanity_3]

No, sff pc enthusiasts know a simple rule of thumb: you don’t want to feed hot air from the lower TDP component into the higher TDP component. Hence, CPU AIO can only be exhaust

[@protator]

With a front rad just make sure you've got more exhaust than intake. Just like Jay said, with a neg. pressure setup you'll pull additional air in through the bottom of the case. The increase in case temp will be negligible. If 2 or 3°C make the difference between your system being 'stable' and board components overheating and your gpu throttling, you have other problems to solve than the position of your AiO ^^

[@fvdeddrift]

The way I had an Antec dual chamber case setup was with a tiny bit of negative pressure, and it could easily be cleaned every other month... I switched to 140's and did some rearranging of the rads... It's been about 2 months and there is next to no dust on the blades, and looking through the fans at the rad fins, and they too are "unusually" clean! Lol! I swear an open test bench stayed cleaner! This is a great post here! Great advice you have provided!

[@msromike123]

And yet people will argue for weeks on the best paste and how to apply it. And watch dozens of hours about which air cooler is best (0.8 F, OMG I need a new cooler) It all adds up man. I'm not about flushing money down the toilet, but if moving the AIO rad makes your GPU and SSD run cooler, it almost seems like a no brainer. Unless you are doing OC competitively the CPU is never an issue, ever. It is always storage and GPU temps that are a problem.

[@joee7452]

The question is not really about stability because the heat difference is not that much. The question comes down to the GPU fans. Not all vendors and fan solutions are the best quality. If the 4 degree difference causes the GPU to have to run 500 or 600 RPM higher, then you get into the how annoying the sound difference would be. Nothing that would stop anything from running, but depending on the card it could turn a barely auditable fan hum to an annoying reving really quick. It really depends on the card and the person and if the decibel change would bother them. With a case it's easy, switch to 140s, switch to silent wings or Noctua, add a fan, there are many ways to deal with heat quietly. With GPUs your choices are much more limited and tend to be costly if sound is the issue..

[@Syphirioth]

Or have an H5 Elite case with fan that pulls air from bottom already to blow directly on gpu XD

[@SIGNALacquired]

the radiator would be best served by placing it OUTSIDE your case, so the heat can radiate away in the open air.

[@JohnSmith-b6n]

I did that a couple years ago with an I9 office prebuilt I janked together. No matter the test, the 10900k would not go above 65-75c in any stress test with a 280 aio

[@MHWGamer]

it is funny how true this is but an entire industry doesn't care for it because it would look like a heater.. or in other words a radiator😂 No, let's put the heat exchanger in a metal body and force the heat flow through tiny gaps while parts of the rad are completly blocked which would have done the "work" for free

[@IS-]

I have never watercooled, but my logic states the same. However, facts have no place in heated arguments, only egos :D

[@MrMcbear]

You would think, but the purpose of a radiator is to have pressurized air move through it to move heat better than ambient air.

[@phunkydroid]

It would also be a lot easier to clean out the dust.

[@---GOD---]

It's all a matter of whether you want to prioritize your GPU or CPU. Top = you want your GPU to get the best cooling by sacrificing some CPU temps. Front = you want your CPU to get the best cooling by sacrificing some GPU temps.

[@Darker_Waters]

The problem is that he didn't test the most commonly used scenario for the argument. Which is gaming and full system use. The result of that testing would be: Top: CPU temps will get a lot hotter because of the graphics card. Front: No real difference in GPU temps, but much better for the CPU. = Front is always better.

[@DM-eq4rs]

Your comment deserves more likes. Straight to the point.👏🎩

[@PyromancerRift]

​@@Darker_Watersdepends on the CPU. A 300w inter trash will make your 300w GPU sweat bullets.

[@PyromancerRift]

Depends on the case too. A good case will have GPU intake at the bot and a PSU compartment behind the motherboard. I personnally use an asus prime case. The CPU is air cooled with fans reversed pulling fresh air from behind and its way better than a 360 AIO on top, over a 7900XTX...

[@extracoolboy]

Very true

[@MO-ss7qt]

"You don't have to fret so much about whether your radiator is in the front or your radiator is on the top." Thanks Jay. This is really the only conclusion you can come to. So many tests have been conducted on this and everything I've seen results where either the radiator on top or radiator in front has narrowly edged out the other. And those tests were no more conclusive for ALL configurations than your tests. Generally speaking, both will work just fine. My mentality is get the liquid cooled with fresh air. Get the air cooled with fresh air. So I like radiator on front. And under the GPU, 3 intake fans feeding it fresh air.

[@CGGeek]

Great video Jay! The one aspect that wasn't mentioned is - While doing CPU + GPU intensive tasks (like gaming) The GPU will be kicking off a lot of heat that will be exhausted somewhat through the rad when its mounted on top. Meaning the warmer air would result in even worse cooling in this configuration during a full PC workload.

[@Nayah9]

• Firstly, CPUs have a much higher TDP than GPUs so they can run hotter. • Secondly, CPUs are far from maxed out while gaming. GPUs on the other hand are fully loaded.

[@prunn]

my rear top would become burning hot with the GPU under load and the radiator blocking the top exaust, I can't see this being good long term. the FE card in this video has a rear exaust which would fix this problem

[@xRaptorScreamx]

@@Nayah9 dude, my 7800x3D uses like 30-40W TOPS while gaming in 1440p, while the 6800x5 pulls like 200-250W, Dont think all CPUs use power like Intel to fuel a dying star

[@Nayah9]

@@xRaptorScreamx I'm talking about temperature, not power... Even X3D chips get hot despite not drawing a lot of power and that's perfectly fine.

[@_fr4mes743]

@@xRaptorScreamx Bro you literally didn't disagree with him. He said CPUs are far from maxed out and CAN run hotter. You literally proved his point

[@BobBobson]

For some of us it's "wherever you can find space." I repurposed a Dell T7500 case and the only place the rad would fit is the removable side panel. Had to dremel out an opening, then drill out holes to mount it. It works beautifully though. Edit: In terms of other component cooling, the side mount rad doesn't change much. The T7500 is by far the highest airflow case I've ever used, and the big dumb front fans move plenty of air to keep everything nice and cool.

[@astanisystems]

Nice. I love PC case modding. I'm currently modifying my BQ Pure Base 500 FX case. Nothing wrong with the case in stock configuration, it's just me chasing more performance and better esthetics.

[@Wooskii1]

I have a Corsair Air 540 Carbide and taking a Dremel to the drive sleds in the bottom and putting 2 140mm intakes in their place actually made a huge difference. I was thinking about trying to stuff my rad in the second chamber but I realized it would be a huge pain. I can't figure out why that's never a factory option, I mean there's pretty much nothing but fresh air back there.

[@06stangsc]

"for some of us we're too poor to do it right and I have to let you know how poor I am"

[@doesitgame]

Those big dumb front fans are massive 140s by Delta aren’t they? The T5500 has similar fans (the same?) and we’ve used one as a server for years, with like two days total system-off time over the last seven years. System hasn’t missed a beat.

[@travissnider5321]

No trash talk necessary, both have bennifits. Top mounted is ideal for asetek design pumps with the pump located on the cpu block , it prevents cavitation, and can help maintain positive airflow with the exhaust having some obstruction. Front mounted gets air directly from outside the case so no residual heat from other components, however it can warm air entering your system providing less cool air for other components.

[@taylorhickman84]

Sounds about right when I had my aio(pump on the block style) front mounted, It sometimes sounded like the pump was loosing prime or cavitating, now that it's top mounted it has been silent ever since.

[@hacoberthejacober3345]

Obviously outside the case is the best option.

[@shpeeva]

except its not pretty

I have one on the back of the case because there was no space inside the case. Doesn't look to bad.

[@ROMANERUIZA]

This is only true if the room where the said watercooled computer is in is either air-conditioned or air-circulated. Had an Alphacool Eiswand 360 running on a FX-9590 in a closed room with a single window and no A/C, and after some time its cooling performance would be no different than an internal AIO's.

[@hw2508]

Outside the house

[@veedubfreak]

Mine is in the window. MORA supremacy.

[@ertugkaya919]

Running an SI for over 20+ years, I loved your response. Thanks.

[@praetorxyn]

Top, and I don’t particularly care thermally. It’s less in the way and looks better. Thermally, could go either way, but exhausting the hot air from the radiator makes sense to me.

[@timmypowers5926]

Yea but it also blow hot air through the rad instead of cool but really don't think it makes much of a difference

[@NALOW84]

@@timmypowers5926 Its generally not a huge difference. Running a front radiator as intake keeps the CPU slightly cooler but tends to warm up the GPU by a few degrees and top exhaust generally will run slightly hotter CPU temps and slightly lower GPU temp.

[@brentvo6662]

Wouldn’t it be more healthier for your radiator to pull clean air through it instead of pushing hot air through it

[@ОсликИа-я2ы]

@@timmypowers5926 Air cooled GPUs are generally harder to cool. At least my 1070 is running at 79 C while CPU is another 10 C lower. Both on air. Add AIO to the CPU - and you're not going to do anything good buy putting radiator in the front, only choke the GPU even more.

[@praetorxyn]

@@brentvo6662 Healthier how? It’s just metal. Its job is to cool the CPU. The heat from the CPU is exchanged into the radiator and needs to be exhausted out, so it’s either exhaust it out into the room, or exhaust it out into your case.

[@gaptoofgranny]

The best configuration for your watercooling system is... outside of the component area. Like in a second chamber. Or outside of the case.

[@airforcedude08]

Trash talking shall commence!

[@trackeduser2577]

my last trash day my barrels were too full to take all the bags out.

[@spencethekeepgaming5795]

bro wants the trash talk because of the SW Outlaws vid lmao

[@jabronilifestyle]

Every time I pass by this one area of town there's a big ditch with a bunch of trash cans piled in it. I'm not really sure why they're just sitting there, it's weird.

[@frankielight2696]

😂

[@forgotten2249]

I got two bags worth out of my car, Bit of a disaster piece but still plenty of trash to go.

[@dyson9422]

i have done it both ways. With the top mount, I had the back two fans exhausting and the front fan as intake to maintain positive pressure inside the case. This PC was used for heaver work loads that were more demanding then gaming loads. With the front AIO as intake mount the loads were light so I used a AIO to make it super quiet.

[@Exalerion]

What I am far more interested in seeing is how front or top mounted CPU rads are affected WITH a GPU under full load, which is a far more common situation for most. I want to know the magnitude of the temp delta that rad placement has on BOTH the CPU and GPU with either and both under full load.

[@msromike123]

Well, we kind of know. The SSD and GPU will run hotter with the AIO rad in front. I suppose it matters on what temps you are already seeing on your system. Are you running an RTX A2000 or a RTX 4090. Big difference between maxed out at 75w and maxed out at 450w.

[@Darker_Waters]

@@msromike123 The difference is that the GPU might be a minuscule amount hotter when the rad is front mounted, but the CPU would become a lot hotter with the rad top mounted. The GPU would make everything else hotter regardless and graphics card coolers are so over-engineered that it wont make much difference.

[@madmax-ox7dz]

Exactly.. sorry to say it but running an occt test to test this topic is just dumb. Like his brain was not on when he made this video or something.

[@Exalerion]

@@msromike123 Obviously, but that's exactly what Jay demonstrated in this video... I'm talking the other way around. Effect of all other components on the CPU with the rad in front/top.

[@DieselThunderAviation]

I have a near identical setup to what you have here in the video. 7950X, 3090 Ti, in a Corsair 5000D Airflow case. Main cooling is a Corsair 360 AIO that is top mounted. No complaints about the performance of the machine, it's quite a performer for me and have never dealt with thermal issues with it.

[@Dtr146]

I've always practiced putting my AIO at the top of the case, mainly because all of the air will get trapped at the top of the radiator.

[@dezpotizmOFheaven]

I have a Motherboard that has two spots to connect external temperature sensors to it. I placed one of those right behind my intake fans and one below the very rear point of my exhaust fans that is furthest back in my case, pushing air outwards through my radiator. I have a push-pull configuration 360mm radiator and 3 x 140 mm + 3 x 120 mm intake with the 120s comming from the side but only kicking in when my coolant goes above 35°C and even than going MUCH slowed than the fans on the front. - The side intake doesn't change much in temps, but I want to keep it an overpressure at all time. The push-pull for the radiator actually dropped temps a little lower than the push only configuration, because the finns on the radiator are infact a resistance for air and can cause air to get trapped and even shoot back into the case. - Adding pulling fans on top helps moving all the air through. I never tried a pull only config. The difference between my intake air and the exhaust (before the fan on the radiator) btw is just 2°C. Therefore it's just another reason to mount the radiator on top of the case to elimintae the chance of getting the burble noises inside the radiator EVER.

[@reidsampey2045]

Jay, you are exactly right on the fact that "cold" actually does not exist. All temps are measurement of heat. My HS Physics teacher would be proud!

[@Martin52863]

So when it is cold outside do I now have to say that it’s less hot outside or not hot outside?

[@decoy1312]

​@@Martin52863less hot than indoors

[@PyromancerRift]

Even though every temp above absolute zero have evergy, we are humans and our body works at 36.6c. This is our zero. Under it it is cold. Over it it is hot. You probably know why.

[@MrSecurity.]

I would have to argue that saying cold doesn't exist. It's only heat is dumb. Sounds like somebody just made that up and decided to keep rolling with it. I mean the definition of heat is the quality of being hot or high temperature. Whereas the definition of cold is of or at low or relatively low temperature or lacking warmth. So it's literally saying it lacks heat. Because heat is hot and cold is not. Leave it to a physics teacher to be all complicated and dumb.

[@carlospuig15]

@@Martin52863 what you actually experience with "cold" is the heat of your body being transfer to a less heat place (air, liquid or a solid object). This is why you feel "cold"

[@PDLaronde]

Jay I find this topic very interesting because its kinda what I do for a living but on a much more massive scale. The removal of heat like this is technically exactly like an ice hockey rink aside from the whole glycol and anmonia system. Heat goes to cold, there is no adding cold, its all about removing heat and regulating it. ❤ Perfectly said.

[@tzfardaya]

Or, hear me out... Front AND Top.

[@mikezappulla4092]

I have my gpu in front and cpu in top. I have a 240 in front, room for a 360 so on the top I put a noctua to try to blow cool air toward the cpu rad. I have a 10900k and 4090. My cpu rarely gets over 70c (usually when using cinebench or when shader cache is being generated) and gpu rarely over 50-55c. I’ve always wondered how the ram and such do but have never had issues so idc.

[@baldcupcake3354]

yep, i put an AIO on my CPU AND GPU.

[@marsMayflower]

I'm listening!

[@chrisk3127]

@@mikezappulla4092 doesn't the 4090 get pretty bottlenecked by the 10900K lol

[@-Gorby-]

Kinky!

[@chienforcer]

I always run rad on top pushing out. Hot air rises (it is why hot air balloons are a thing) and heat moves towards absence of heat (see convection) which means in the top position the air is already moving in that direction. So it always just made more sense to me to cater to the natural properties of heat radiation.

[@AdmrlSpeedy]

I solved this conundrum by having radiators everywhere.

[@Darker_Waters]

Or none. ¯\(°_o)/¯

[@jensv874]

Or do the opposite like me and have no rads and enjoy a silent pc.

[@j03_mc]

​@@jensv874water cooling makes it silent

[@Stevo-LSD-25-mk4fj]

Hi Jay, I’ve been a long-time fan of your content and appreciate how you always advocate for better consumer experiences in the tech world. I wanted to bring something to your attention regarding Corsair's premium mice. As someone who expects a certain level of durability from higher-priced products, I’ve been disappointed with how quickly their mice, particularly the mouse wheels, start to fail. For a premium product, I feel the quality isn't holding up to the cost. I've seen several people mention similar issues in forums, and I believe your platform and voice could help shed light on this. It would be great if you could dig into this and possibly encourage Corsair to address these issues or improve the durability of their products. Thank you for your time, and keep up the amazing work! Best regards, [Steven}]

[@Stevo-LSD-25-mk4fj]

Just wanted to add. That besides their mice, I've also had keys break on their K95 platinum 🤬

[@deadghost1189]

Both are fine. Looks better at the top.

[@jabronilifestyle]

So you're saying you don't like the front?!? (women logic)

[@finnguy1549]

@@jabronilifestylewhat??

[@geckoop2848]

@@jabronilifestyle🤦🏻

[@jabronilifestyle]

@@finnguy1549 dw you'll get a girlfriend some day.

[@SafetySheepRnD]

So my theory for why the temp curves looked they way they did is because the heat generated by the test is causing convection currents inside the case which causes the early temp drop. Before the test starts, the case temps are mostly normalized with the environment, but as soon as the heat gets generated at the intake, the heat is injected into the case warming up the air that rises and pulled fresh air from bottom intake. Overall, the difference was a lot more significant than I expected. Great video

[@TKevitt33]

So minor correction here but “closed systems” are always trying to reach thermal equilibrium. That is to say in the presence of heat all things will heat until the same temperature is reached and in the absence of heat all things will cool until the same temperature is achieved. A difference in temperature in a closed loop system mean you’re imparting some form of energy into the system.

[@_fr4mes743]

Just a reminder to everyone else that a pc is NOT a closed system.

[@alaingraham]

Basically; the further from ambient temperature, the more work is required to maintain a temperature difference higher or lower in a way that drastically increases for each degree above or below. The more work, the more electricity it takes, and if you increase ambient by 5c then it's going to increase the amount of work required to reduce the temperature another 5c from ambient to the same temperature as before, and if you've already saturated the devices cooling capacity, then it's gonna increase the temperature across the board by 5c.

[@alaingraham]

And that is the physics behind it. I think, it's been a while, since I was taught this at secondary school.

[@Fmily]

Pre video trash talk: The pump needs to be the lowest part of the loop. Doesn't matter if it's an AIO or a custom loop, the pump needs to be the lowest part of the loop.

[@Detroittruckdoctor55]

Thank you, jay. I'm building my hotrod 1000D obsidian case, got prolly 4000 dollars in parts waiting to go together, in my first custom water loop build. I Unfortunately went with the fancy Corsair QX series because I thought it would help me with my poor cable management skills. I'm now finding out they might not be the best performing cooling option for radiators. I almost think you were lurking in my AMD pc builder group I've been hanging around in on Facebook. I'll keep lurking on your custom videos because I don't think the guy at Micro Center was into my custom loop when I came in there to buy my radiators and pick his brain. Also me and my kid brother, who's fighting leukemia and in chemo treatment, is wondering how your cancer journey is going. I put your videos in the background to distract us from the health issues. Me and gaberial will be rooting for you. PS ex girlfriends brother , who I still enjoy talking to, said, "Who do you think you are attempting a custom water loop. jayz 2 cents?"

[@jodythornhill]

The answer to this is easy…yes!

[@BlueFlash25]

But the question was … oooh I see what you did there 😉

[@Taliseth]

Love those people that say you'll cause thermal throttling, obviously if that were the case no matter what config of the rad, it would thermal throttle or you did something really stupid like block all vents, and even then you'd probably be fine still and the cause would lie elsewhere. On that note, most people jam their pc in the worst location anyway for Thermal throttle by "heat soak" to be a case due to a front mounted intake rad. Would love more videos like this though. It would have been nice to have seen the other fans set to auto as well just to proof your point or to get a comparison. A video on the correlation between ambient and component temperatures would be awesome to see too. I run my rad at the top as intake, which goes against heat flow mechnaics, but it works good enough Front of the case is to far away all components still sit around 30-40 idle and 60-70 when stressed 100% so happy with that. Could i have lowered the overall temps a little by properly, would it have made a difference no, rather have the hardest to cool component (CPU) be slightly cooler and rest slightly hotter.

[@AudioAdvisory38]

But what if it is on the side!?! *screams and jumps out window having learned the PTA is disbanding (Simpsons)

[@magnusnilsson9792]

I have my 280 RAD on the top right hand side in my Core V21, and my 2x120 intake fans feeding the GPU first from the left side. In addition to a 200 intake in the front and a 140 exhaust in the rear.

[@kevinvandenbergh9998]

ive been eyeing new cases exactly for that reason. Id want to put my rad on the side, so there is some fresh air coming through it (or slightly warmed) but then it gets blown away by the wind coming from the "empty" front fans then have top and back as exhaust

[@Alovon]

Assuming a standard layout case with equal Rad space for front and top. Probably top becuase 1. Space, it's the most convenient and least in the way usually. 2. Airflow for the GPU. It's easier to intake cool air from the side than it is the top (hot air rises and all). So having intakes on the side/Bottom and exhausting out the top is the most balanced in my mind for most case layouts. 3. Unless you're cooling an Intel i7/i9, most CPUs nowadays, and especially the AMD X3D SKUs for Ryzen 7000 sip enough power to get away with a 120mm Rad. So a 360mm rad anywhere is a bit overkill for the CPU itself, and it actually can more serve as exhaust for the GPU at a certain point. So I just built my new PC with a 7800X3D+4080 Super config in the Phanteks NV5. So I have all the fan slots filled to give the case the intended "Diagonal" airflow config (air intakes from the side, pushed up by the bottom intake, then pulled up and to the side by the top and rear exhausts). Becuase I'm using a 7800X3D, my 360mm rad is mainly there to fill all the fan slots on the top of the case in order to push more air out for my GPU without my CPU recycling heat into the case itself.

[@philwithcheese]

GPU puts way more heat into the case than a CPU. The real reason to front mount us because the CPU will heat up quite a bit more than 4C when using a too mount and running a GPU load.

[@SagolSpams]

The timing of this video was amazing, and it came out a couple of hours before I started my new Corsair 4000D build with a Corsair H150i LCD AIO. I was so torn on front mounting that AIO on the 4000D, but all has been going well! Clutch timing with your video on the core parking with the 7950x3d too! Thank you Jay!

[@OGGlasses]

0:23 i honestly would think front mount would be better as its pulling in fresh air and not warm air from case

[@florianhug9801]

Then on the other hand you gpu gets more warm air. So you simoly have to get your priorities

[@rubnim4870]

I totally agree. Just keep in mind the layout of your PC case. The configuration could have more influence on your personal setup, if the top has a limited amount of hot air to get removed (size of the mash, area of the mesh, glass plate etc). The same should be checked for the front side. It's not affecting Jay's testing, but check yours. My decision: 😅 rads on the front an top, self build loop. The front rad is just 30mm thick, the top one is 45mm thick to exhaust more hot air from the case and the front one doesn't add that much of resistance to the case air flow.

[@BigDrewski1000]

So lesson to be learned? Fan orientation makes more of a difference than radiator orientation. I imagine if you had had a couple fans in the front to bring in air with that top mounted test you'd probably cool down those extra degrees that you had with the front mount

[@ronijs19]

He had. 6 fans, only rad moved. 2:52

[@modru2004]

or if he had ran them at full speed instead of capping them. he starved the rad for air intentionally in his top mount setup.

[@BigDrewski1000]

@@ronijs19 Right, but what I was saying g is I feel like he would've gained those few degrees of cooler temps if he had had fans on the front along with the fans on the rad up top

[@BigDrewski1000]

@@modru2004 he just ran em at stock speed, right?

[@modru2004]

@@BigDrewski1000 he ran the front fans at 65% when using the top mount setup.

[@michaelsumnerjr5819]

Love you, first of all. Second of all; you have to choose so many things when it comes to each individual build. But I will admit I tend to have that min/max type thing in me, and when I have debates and discussions about thermals, it is very funny how some people are like "well it's only 7° difference" And I'm like what!? "only!?" ...that's huge! But in this scenario it's interesting to show that for the CPU, Front intake is better, btw a Push/pull might reduce the temps another 1-2­° (so might net more like 4°). And the CPU is the most difficult component to cool out of all of them where ... dare I say, the temperature really matters. Also, this is just one thing you can do. I keep saying it let's say you also use Liquid metal as a thermal paste, BAM! there's another 4° if it was an intel i9 14900, use a contact frame there's another 1-2°, more fans in the case with neutral or positive pressure, there's another few degrees, it's all the little things you do, that adds up to a lot.

[@Azmodon]

So that's why it worked... Built a buddy's computer earlier this year in an black O11, 360 on top, with three 140's on the side, 3d printed a white duct with acrylic windows (storm trooper rave box) from the top most 140 that stretched out to feed the entire rad. Had about 1/8" gap between the duct and rad fans, with an open rear so this "1-in 3-out" setup was pulling the excess from directly above the ram and vrm (case air has to flow against them to make it out). Also had the 2 bottom 140's ducted into a 7900XT (duct mounted to PCI slot covers), with a 1/4 volume diverter to the case, + the normal case exhaust fan. Technically a 4 in, 4 out, with 5% negative cfm.

[@meibing4912]

The temp difference is unlikely due to the placement of the radiator. The radiator will move the amount of air it can move regardless of where it is and that air can only come from the outside. However, hot air "escapes" much easier through the top when using a front radiator thus assisting the fans in getting rid of hot air faster. A top mounted system will let less hot air escape through the front (in addition to the radiator fans themselves). This is because air is very difficult to compress (so we could also install the fan in the middle of the chassis and get more or less the same results. And...yes: we actually do exactly this when mounting a fixed fan! It works great at sucking in new air and thus pushing out hot stuff). Chassis format and grid layout is thus hugely more important than radiator placement. Have mine on top - maybe I could get a marginal gain at the front. Could not be bothered. YMMV.

[@BlacKSye]

Instructions unclear. Trash-talked throughout entire video and got weiner stuck in a toaster. Please advise.

[@Jeff.78]

Sounds like a Marine tried to make a hotdog🤣 Unplug toaster. Apply butter. Twist toaster slowly. Gently pull out grill marked sausage. Pay me $420.69 for the help. Have some liquor and crayons. Go to bed. Wake up. Do not repeat the prior days events.

[@Masaliantiikeri]

In this house we obey the laws of thermodynamics! -Homer Jay Simpson

[@hogibunz834]

Haven’t watched the video yet and I’m fairly new to pc’s. My first thought is that using the radiator on the top with exhaust fans will be the most efficient way to keep your temps down AS LONG AS you have enough intake fans at the front or bottom to keep the pressure in the case positive or neutral. Using a radiator on the front with intake fans means the hot air will be entering your case and the rest of the components will become more heat soaked and less efficient than if the air was exhausted out of the case after the radiator. But like I said, this is only a good idea imo if you can have enough intake to keep positive pressure in the case. I think that the cooling the cpu will be more effective when you use the rad in the front, but cause more heat soak to the rest of the components inside. It’s a give and take. Do you want more performance at the cost of more heat for the rest of the components, or cooler temps for the rest of the components, but not as much performance from the CPU.

[@farmerowga5890]

Sorry, but this wasn't a good video. You should have tested also with GPU load. You should have said that "Setup 1" is better for CPU heavy tasks (professional work) and "Setup 2) is more optimal for GPU heavy gaming. You should have spoken more about possitive and negative pressure affecting the tests.

[@scottperry5454]

This... My 4090 cooks the inside air temp when running. Without a gpu load, it's just circulating outside air.

[@RandomRobbert]

I like the testing you did, but I am missing some worse/IRL data results. (This is meant as friendly feedback.) Meaning: gaming load fan speeds at 30% and/or 50% fixed, maybe silent freak 20%. I think it would be most interesting to plot the data to see the delta temperatures for: - Delta Air intake case -> Air exhaust case (should change depending on total airflow through the case) to test airflow impact - Delta Air intake rad -> Air exhaust rad (should change depending on placement in front or behind GPU, also CPU load/fan speed correlation) - Delta Air intake GPU -> Air exhaust GPU (same as rad but the other way around) - Delta vs. open test bench to see zeroed-out difference for all scenarios. Essentially, (grow long hair and a majestic beard and sell a copper cup and coasters🤪) If you fly me to you or ship the parts, I will happily do the testing for you 🤣

[@Aina_Nightfall]

For me here are only two solutions with cat in the front, without cat top mounted

[@Tiki832]

Weird timing for this video as I was recently discussing this very thing just 2 days ago, to say it doesn't really matter I think is a bit of a gloss over summary though. The take away I feel is that you should consider the two options as being significantly different but one with a difference that may or may not have any real significance depending on specific situations. What is the actual size of the case? To say there isn't a difference in a large case chassis where there's a lot of room for warmer air to be pulled in from the front into the case and then exhausted from the top is entirely different to suggesting there's would be no difference when you're doing the same with a MicroATX form chassis. How long is the system going to be switched on? If you're going to be regularly turning off the machine or if you intend to essentially keep the system on running 24/7 can also play a part in which configuration is more suitable. And various other variables that can mean the actual impact between the two can be much more significant depending on the finer points of that particular build. When I first got my machine, I went with the Corsair 500D case with a Corsair Hydro 100i V2 AIO. Went with a pre-built machine as it was a Ryzen 2000 pre-order machine that offered day 1 of release dispatch of the processors rather than having to wait for release of the processors and then buying everything separately and waiting a few days after release for dispatch. The engineers building the machine initially went with a front pull through the AIO and top exhaust through 3x fans configuration, as I knew I tend to leave my systems operating 24/7 and that the machine would be in a room that will often be warm and stuffy (The joys of being a Brit in the UK where houses are designed to retain as much heat as possible and air conditioning isn't really a thing), pulling often very hot and stuffy ambient room temperature through the radiator stack that would be getting next to zero system off downtime wasn't really something I liked the sound of, so I asked if they could switch it around and have the 3 additional fans I had ordered with the machine be placed to pull from the front and then the AIO radiator stack be set to exhaust from the top, which they did for me without any complaints or resistance. As I went with such a beefy case as the 500D, honestly it likely wouldn't have made a whole lot of difference, there's enough room in the case for air to flow nicely and for various components not to suffer too much from having the AIO radiator stack pulling already overly warm ambient temp air through the radiator that's been operating 24/7 for 3+ weeks. However in a more smaller frame chassis, I'm pretty convinced the difference in that particular use case, would have actually been much more significant. So when watching this video I hope peoples take away isn't that "It doesn't really matter", but rather that the two different configurations introduce different sets of situations that will can have more of a impact on certain build environments, and you need to be taking that into account especially when you're taking chassis form factor into account.

[@jabronilifestyle]

12:19 if there are still many many many PC enthusiast groups, why did Anandtech shut down? Kinda weird since there's still clearly a demand for content of that type.

[@Drunken_Horse]

The type of content that was getting produced that wasn't either readily available in quicker to digest formats or was not so niche that few people would want to see it was enough to generate enough revenue to keep the lights on.

[@kentchenery1813]

For the front mount the air in the case passed through the "cooling" of the radiator (as it was pulling it into the case). That is, its heated by the CPU as its entering the case (via the radiator). And so other components are affected by that. When it is in the top the heat from the CPU is dissipated into your room (not the case) and only "cool" air is pulled in with the negative pressure (through whatever path it can take). So yeah... these results make sense. Pushing air out of the case in both approaches will give different results again (probably very very similar temps at a guess). TLDR: I think these results more likely show the benefit of push vs pull more than top vs front.

[@msromike123]

So, putting it on the front just means it will be louder since all the case fans have to work harder to keep the interior cooler. Other than geek benching my CPU it never throttles, not so much for the GPU however. To me this just proves for gaming, you put AIO radiator on top. GPU cooling over CPU cooling all day long. With the bottom fans on a curve tied to GPU temp and the back fan blowing on the SSD on a curve. and the backplane fans blowing in fixed at a rate that is inaudible. Works for me. And Noctua 80s DIY mounted to the 3.5 drive bay grills and blowing into the drive bays. Those exhaust behind the backplane and run on a curve keeping them as close to 39 as possible, at all times (load or no load.) Mechanical drives do better when they are the same temp as much as possible, not so much letting them cool, heat up, cool etc.

[@jprepo1]

This is true in theory, it just doesn't matter in practice outside the hottest and most overclocked setups

[@msromike123]

​@@jprepo1Fair enough but my EVGA 3070ti hits 80 enough and on occasion 85. I would rather not see 84 and 89, like I might if the AIO was in front. Pretty simple, put it where you want. If you decide your GPU runs hotter than you would like, put the rad on top and have good fans blowing on the GPU from the bottom on a curve.

[@Syphirioth]

@@msromike123 Buy elite H5 case. I think that angled fan blowing on bottom of GPU makes lot of difference with fronmt radiator.

[@-Good4Y0u]

Interesting video, I watched this when I first installed my Artic iii 360 aio on my Fractal North. I had to install it in the front of the case with the fans on the front, then the radiator on the inside. I used the Fractal included 2x 140mm fans to the top and used a Noctua F12 120 I had left over for the rear exhaust. I would have been pretty interested in the push configuration here with those default fans. I definitely want to see that video. Also here is the AI summary for your video this time: This video by JayzTwoCents tests the impact of mounting a radiator on the front vs top of a PC case. Here's the breakdown: Test Setup: AMD 7950X CPU Arctic Freezer 360 AlO cooler Antec Flux Pro case RTX 4080 Super GPU (used as a passive thermometer) 1-hour 0CCT AVX2 test with all cores loaded Results: CPU: Front-mounted radiator resulted in a 2.2°C cooler CPU temperature. GPU: Front-mounted radiator resulted in a 4.5°C warmer GPU temperature. Conclusion: While a front-mounted radiator can slightly increase the overall internal case temperature, it's not a significant enough difference to cause throttling of other components. The difference in temperature between the two configurations is negligible and likely won't impact performance significantly. The video debunks the common misconception that front-mounted radiators cause significant heat soak and throttling. Key Takeaway: Don't stress too much about where you mount your radiator. Both front and top mounting have their pros and cons, and the difference in performance is minimal.

[@social3ngin33rin]

Exhaust top mount. Intake front fans, intake bottom fans. Exhaust rear fans.

[@DjTinnio]

Yep.

[@john1182]

that works for basic cooling but what if you have a cpu and a gpu that are on AIO's ? you still intake at the front and exhaust at the top but in my case the cpu is front intake and the gpu is top exhaust as its a good 300 watts of heat and all of that goes out the window if its one loop

[@social3ngin33rin]

@@john1182 stick to air cooling :P

[@SabentHD]

​@john1182 Wouldn't be an AIO if it is full looped lmao Second part, I believe Jay has already made a video on shared vs non shared loops, inwhich the awnser to that is the awnser of this video, which basically changes nothing

[@john1182]

@@SabentHD i was giving 2 different senarios so there is no need for the lmao, im well aware of the outcomes on the different builds as ive been building and reviwing my self for 20+ years. if its seperate loops you put the one with the leess heat on the intake, if its one loop all it does is warm up the ambient air in the case a few degrees, but thats were airflow helps keep componant temp down

[@benoitbonnier7309]

I am not a computer geek like you guy's are, but learning a lot. I'm actually getting ready to build my new computer because my 14 year old one is starting to frustrare me a lot. What I had read is, the case fans already installed should remain is their position, this mean the radiator as to be installed on top for max air circulation. I won't have the top knoch computer but this one should be good to me for another 10 years min. Not a big gamer but more into CAD.

[@ackos2488]

My cat apparently liked to pee in the top fan hole, bricked my desk top.

[@mikesunboxing]

Was it a Corsair case? Our cat liked doing the same thing on a 4000d

[@kevinerbs2778]

Kitty is mad they didn't they attetion they wanted. Did that for spite

[@Tx_Jay_]

​@@mikesunboxing Didn't expect to see you here 😂 but dang that sucks

[@mikesunboxing]

@@ChandlerUSMClol great product name

[@ianollmann9393]

I bet you'll think twice now before focussing back on your game when kitty is hungry!

[@Asiri010]

The reason for the steady increase in GPU temperature during the front mount test is due to the location of GPU temp sensor ( which is supposed to to be around the middle). Hot air coming from the the right end causes the right side of GPU heat sink to warm up gradually. This heat is transferred by conduction until heat finally reaches the temperature sensor which , in turn, will start responding to this heat absorption and start recording a steady increase in GPU temperature. I've tried all configurations with my 13700K and RTX 4080 and believe it or not, the optimal configuration that works for both my CPU/ GPU temps was to top-mount my rad , set the rad fans to intake, modify GPU fan profile a little bit , and set my front fans to exhaust.

[@vsm1456]

that's a bit bizzare, but curious. could it be something about your room? like if top intake gets cooler air than the front one?

[@Asiri010]

@@vsm1456 I believe it’s due to coolant flow being higher when the rad is top mounted. When the pump is at a lower position, the suction presuure will be higher than if the rad is front mounted and that will increase coolant flowrate.

[@iiisaac1312]

As for heating up the insides, I'd be more concerned about the components themselves having a shorter lifespan due to heat. Around the 2000s, Dell's Optiplexes were notorious for cooking the hard drives and caps on their motherboards. Many IT departments had to replace motherboards and HDDs every year per each system because they got so hot inside. The CPUs can handle the heat though as they have built in thermal protection, throttling, and are usually relatively tough (except for high end 13th and 14th gen intel), the weaker parts like the motherboard can't. This is why used CPUs are so cheap but used motherboards are so expensive; the CPUs survive but the motherboards don't.

[@dethriguez]

That’s the year 2000. Where dell optiplex was not cooled with an AIO. It’s 2024 now bud.

[@iiisaac1312]

@@dethriguez They recycled hot air used to cool the CPU from the exhaust back into the intake, acting like a front mounted AIO that blows CPU heated air into the chassis. Also, "It's current year!" isn't an argument.

[@dethriguez]

@@iiisaac1312 Which still isn’t an AIO. Also, it’s almost 2025 bud.

[@iiisaac1312]

@@dethriguez And what do either of those have to do with the long term consequences of heat exposure causing electronic failure?

[@dethriguez]

@@iiisaac1312 and what does a dell from 2000 have to do with an AIO top or front? Nothing.

[@mojojojo6292]

Depends on the PC case as well. I have a Lian-Li O11 Air Mini and the 280 radiator only fit's well in the front and also looks better there. Top mounting with 140 fans blocks the custom cpu power connector I use. I had to put a 120 and a 140 on top to fit it. The O11 has 2 fan mounts on the rear side panel near the front that can be used to exhaust most of the front intakes instantly. Got 2 bottom intakes to feed the gpu then as well. Works great.

[@DeCapitanOG]

5$ on it doesn't matter.

[@andreasschelfhout3472]

In regards to the 1 to 1 increase between ambient and component temperature. Forced convective cooling between a solid (what needs to be cooled) and a fluid (the coolant) is described by Newton's law of cooling: Q = hA∆T where: - Q is the amount of heat being transferred. This needs to equal the amount of heat being generated in order to maintain balance. - h is the heat transfer coefficient, which comprises several physical parameters such as heat capacity and conductivity of the solid and fluid, velocity of the fluid, shape of the area interface, etc. For relatively low temperature differences, this parameter is constant. - A is the surface area of the interface between the solid and fluid, which is also constant. - ∆T is the temperature difference between the solid and fluid. Since h and A are constant, the only way to change the heat transfer is by changing ∆T. Or in other words: in order to transfer the same amount of heat, ∆T needs to be constant. Thus when the ambient temperature goes up, the component temperature also goes up 1 to 1.

[@C-M-E]

The ambient air temp part is actually quite simple. The relationship is linear between temperature going in vs component rise, as the temperature per a volume has already absorbed x amount of heat before getting to your fans. If it's 73*F going in, it won't drop to 71 for instance just by being pushed or pulled from a fan into another volume. Fans aren't responsible (directly) for cooling, it's just moving air past, into or away from another volume. It's a baseline for lack of a better term. Air won't heat soak components so long as it's being exhausted at an efficient rate compared to intake. Ideally, even if you're only running one fan for exhaust, run it at a higher rpm and the air stream will be in constant motion and will actively be looking for other exits to passively leak out of if the exhaust fan isn't moving sufficient flow.

[@Carp_Racing]

I run mine on top, because my case had fans on the front already.. LOL Jay, thanks for this and all the other videos, whenever I have a question I come here first. I can normally count on your channel to have the answer to all my questions.. Thank you

[@petrusiusmaximus]

Nice Video. Some Thaugths: 1.) I would like to see a direct comparison between Push-, Pull- and PushPull- Configurations. 2.) We need better Computercases because it makes no Sense to blow the waste Heat from the Radiators into the Case and, if necessary, if the GPU is also Water-cooled, push it out again through the second Radiator. It would be better if both Radiators radiated the Heat directly Outside. Even better would be a Two-Chamber System where the Radiators are cooled separately from the Motherboard and GPU, such as possible in the ThermalTake TheTower900. Even the slightly smaller ThermalTake The Tower 500 allows the Heat to be dissipated directly from the Computercase via two 360 Radiators via PushPull Configuration and still add cold Air via four 14cm Fans and extract it via up to Three More at the Top of the Lid around the Motherboard and RAM in addition to Cooling. 3.) It would also be ideal if the waste Heat disappeared completely from the Computerrooms and was used via Heat-Exchangers, for example to support the Heating of the House. I already used it to heat my Thousand Liter Aquariums 20 Years ago. At the Moment when these environmental Fanatics are forcing us here in Germany to tear our Oil- and Gas- Heaters out of our Houses and the Energycosts are being driven to insane Heights, it is certainly worth thinking about using this expensively generated heat sensibly, especially if you have a good 2 dozen Computers and runs Servers at Home for Work. Of course you don't have to overdo it like Linus did ... 😉

[@tonynubares6880]

That is why I trusted my own engineering capabilities rather than random guys discussing on forums. I bought an Enthoo Pro II case and got: - a top radiator (coincidentally an Arctic Freezer II 360) in PULL config - 3x140 front fans, pulling as well - 1x140 fan for the exhaust. This is the real benchmark I'd like to see. I can say that my 10900k hasn't melted yet :D

[@theorangecactus2177]

8:52 The reason is because the heat transfer rate (in units of power - watts) is directly proportional to the difference in temperature. If the temperature difference is doubled, the heat transfer rate doubles as long as all other variables are constant. Absolute values are irrelevant here, only the delta matters (it doesn’t matter if it is 100C vs 0C or 500C vs 400C, still a 100C delta either way). Because the heat transfer rate must stay the same at any given workload (in equilibrium), the difference in temperature must also stay the same as all other variables stay constant.

[@Kmmlc]

I think a huge part of this depends on the case in question and if you're using an AIO (like I am) or if you're using a custom loop. Custom loops have so much more flexibility since you can put everything where it's needed without worrying about the pump as much, especially if it's a pump reservoir combo. AIO's on the other hand limit the choices. I had a MSI Core Liquid MAG before I got my Corsair one. The mag had the pump in the radiator and the hoses were right next to the pump. There was always the sound of air going past the pump and eventually it stopped cooling. That was due to the crud buildup in the heatsink part though. But I bought a case big enough to allow me to cool either way, I just prefer top mounted.

[@kfancey1]

Your explanation of thermodynamics is spot on! Simple and effective.

[@ImnotgoingSideways]

Top mount: minutely warmer water. Front mount: minutely warmer case air. Dual radiator: hot exhaust air. In the end 1W = 1W and as long as you don't experience thermal runaway, it's six of one and half dozen of another.

[@Aloysyus]

Two things, i hope i didn't overhear anything you said about that: 1. Most people don't have intake bottom Fans below the GPU as well. How would the results be without those? 2. It rarely happens that both GPU and CPU run into their power limit. When gaming a CPU normally doesn't soak up to 180W or more, in case of a 7800X3D it's barely 60W. So when gaming and there's full load on the GPU - there wouldn't be coming 180W of pre-heated air into the system.

[@Nayah9]

• If absolute silence is what you're after, then radiator placement absolutely matters. • I have a 7950X set to 105W Eco Mode and all of my Noctua case/rad fans are set to a ridiculously low fan curve so they are barely audible at full CPU load. Max CPU temp is ~80C. • The loudest fans in my PC are the ones from my RTX 4090 - MSI Gaming X Trio (set to silent mode and undervolted to keep the same performance or more while drawing ~20% less power). • Gaming X Trio cards don't have a vapor chamber so they run hotter (yet stay very silent). My max GPU temp is ~76C. Having a top-mounted rad allows my GPU to quietly stay below 80C.

[@retiredcryptohunter9031]

The temps are slightly hotter when mounted in the top because the radiator is mounted horizontally so the coolant can get dead spots in the fluid flow at the highest point in the radiator. You can test this theory by doing the same test on your open bench by mounting the AIO vertical and running the test and then horizontal and running the test again and I am sure you will see very similar results.

[@moonchild666]

Great video. Bear in mind that most realworld scenarios would involve a GPU spewing huge amounts of heat into the case. A top mounted rad would have this hot air pushed/pulled through it. Not the greatest situation for efficient CPU cooling. This debate will run and run and the correct orientation will be governed by what is being demanded of the whole system. Gaming? Video editing? Depends what you're doing.

[@YOEL_44]

I would still try to put the rad in the top, that way I can have my fresh intake fans running slightly faster to generate positive pressure and keep the system cleaner.

[@everestcomputer]

ive never seen people argue about the heat difference, only the gurgling and possible pump life difference due to it being front mounted with tubes at the top...

[@jabronilifestyle]

I purposefully killed a Corsair AIO by ensuring that the pump was above the rest of the unit because I wanted to see just how delicate the pump was. It died real fast. Now if someone tells me improper placement of the rad won't kill their pump I can confidently tell them that they are simply wrong.

[@Syphirioth]

@@jabronilifestyle I say front placement. With connections on bottom. Then the air bubble will be in radiator top. And there should be big margin before that water level in radiator gonne be lower than pump. And should keep tubes and pump always filled.

[@MeinaAI]

Thank you very much for this video, i lost my mind with how many times i had to read that my gpu radiator was wrong by being on top and what not. The reason i had them on top is because i assume throwing the 600W of heat a 4090 generates would heat up every single thing inside of my case instead of just throwing it out through the top.

[@crypto_que]

I was wrong, I previously thought that all AIO should be on top set as exhaust because of thermodynamics. However I was wrong. I saw a test where it was explained that pulling hot air out in this manner ensures that heat is inevitably circulated through the cooler and back to the CPU. AIO should be set as intake whether top or side mounted. It’s more important to get as much cool air into the case vs where the radiator is mounted.

[@MrSecurity.]

I know you live in California but setting your thermostat to 73°, is crazy. Room is above 70. It's too damn hot. I live in South Dakota, particularly the Black hills so in the summer it can be 100° and in the winter it can be -30°. My AC is going 24/7 all summer. I like my room to be at least 68° Max in the summer, preferably like 65. In the winter I leave my air conditioner in the window and I usually put duct tape over my heating vent in my bedroom because I like my room to be like a minimum of 50° in the winter because the cold helps me sleep. 50 does get pretty cold sometimes. That's why I said minimum. If I can have it at least 60, it's perfect. Sometimes that doesn't go to plan. It gets really cold and my room has been down to like 42 before. That's pretty cold. I took the duct tape off the vent on those days. But when it's like 50° in my room, my PC does stay pretty cool. I should also mention that my house was built in the '60s so it doesn't have any insulation in the walls. But back to my original point. How are you comfortable at 73°. I know it's like 85 to 100° in California, but damn. 73 is too hot. I mean I'm a big dude so I know I'm running a little hotter than most people but still. What makes me mad is most air conditioners only go down to like 60°. I went to Texas one summer and the central air unit in the timeshare I was staying at went down to 50. That was awesome. Texas is hot as balls. It's also hot and humid. I lived on the other side of South Dakota for 5 years in the plains and it's 110° everyday of the summer and it's like 100% humidity. That shit sucks. At least on the west side of the state when it gets hot. It's fairly dry

[@iamgadgets]

Excellent video, bro. Personally I think negative pressure is the way to go. As I understand it negative pressure is a partial vacuum, and a vacuum is technically a heat sync. Vacuum wants to absorb heat, among other things.

[@albertmas3752]

Well. All before watching, the basic theory would be that as hot air tends to go up so it should exit the system faster. That added to the point that a front mount means that the hot air goint up in the lower parts of the radiator when hot air is exhausted would heat up (very sligthly but it's there) the higher parts of the radiator. This from a very basic perspective. There may be a thousand things that might change the outcome but regarding the most basic o basics, it should be like I said. But to make the test fair both positions sould be as intake or exhaust. Because using on top you're using air much hotter to cool down the rad than the front position and intake where you're using much fresh air. A you said in equal conditions difference between push or pull in negligible but the temp of the air used to cool the radiator is affects in a big way the final temp the cooling device can reach.

[@pr0xZen]

Personal view/perspective: If you're running a somewhat recent, higher end aircooled GPU, do CPU (AIO) rad _intake._ Don't drag all that hot air from a 300W++ GPU through your CPU-cooling rad, just get it out ASAP without impedance. If you're watercooling your GPU too - don't neglect general case airflow. With serious power beast GPUs, CPUs, possibly early days pcie 5.0 NVMe drives, high spec DDR5 - your motherboard chipset, VRMs, high performance I/O - all that stuff needs some airflow for cooling too. Even with premium-stuff custom loops, a lot of people still suffer performance issues because they get too hyperfocused on dumping the heat off the things with a waterblock on it. And by neglect or mistake end up with everything else stuck in airflow dead zones, just soaking in their own heat and what radiates off the CPU and GPU blocks.

[@space.raider.2271]

I enjoy using 420mm radiators which can't be top mounted in 99% of all cases. All you need is a quality pump and a properly topped up loop with minimal air content (which you will get in an AIO unless you go cheap) for a front mount to perform perfectly without any additional noise.

[@DoNotFitInACivic]

Custom loop here. 480 top rad, 420 front. Both with pushpull fans Waterflow is CPU, top rad, front rad, pump/reservoir. I also am still using my old Corsair Dominator DRAM fan. Everyone stays below Sixty-Five. Which considering the $$$ i dumped into that loop it better be cool. Lol

[@WardyLion]

After watching yours and Steve's last videos on the issue, I went with top mount and, if the numbers being reported are anything to go by, even with the AIO fans pulling air from inside the case, things remain well within acceptable temps.

[@Deeppurp]

Comment during adblock: While there maybe some variance between front and top, the most important mount is making sure the pump is not in the highest part of the loop. Easiest way to make sure: Rad on top Second easiest: Rad in front, hoses and top of rad mounted higher than pump.

[@davidhines7592]

this makes a lot of sense. less than 5 degrees sounds about right. i bet these people dont stop to think your television, and other devices, and even your lights generate heat. i need to know this stuff because i have a medical condition meaning i cant hydrate as easily and get dehydrated much easier in heatwaves. i found by having a full thermal blackout curtain, blocking all solar gain from coming in, plus leaving off the tv and lights, and just having the monitor and pc on, i can keep the inside temperature 6 degrees lower than it would have been, usually its at least 8 degrees lower than outside in the shade. this is how i stay out of hospital attached to a drip to rehydrate me in summer now that global warming has changed our summers here. all those will affect the temps too, as well as aircon if in america. ideally you would now find some test that stresses only the gpu and not the cpu and see how that affects temps, then use both results to predict an approx. temp for a combined test... then do the combined stress test. i dont think anyone could really argue with that done.

[@dien00b31]

I have a front and top Radiator (both GPU und CPU are water cooled) and I have absolutely no temperature problems. Front is push/pull and top is push. The rear fan is the only fan set to a static speed, making sure my airflow puts the case under a certain pressure to prevent dust and heat buildup. Works perfectly fine.

[@isturma]

Pre-Video Trash Talk: Mine's in the top, so I didn't screw up the hose placement. Post-video trash talk: My case ALSO has fans mounting locations under the GPU that bring in fresh air for the card; but I let mine run because... well, it's a 2060KO OC and the cooler could've been better designed. My card runs hot (doesn't really overheat) but I transplanted my components to another case where I could bring in fresh air to help keep it in check. Cheaper than buying a new card. I really want to say something snarky but i'm not really a sociopath.

[@mdb4879]

What I'd like to know is if you're water cooling your CPU and GPU, which do you prioritize where? CPU radiator in front and GPU on top? Or vice versa? Does it depend on application? (Gaming/GPU heavy tasks put the GPU radiator in front, CPU heavy use put CPU radiator in front?)

[@oswaldjh]

I tend to use the top config with an additional fan on top and a rear fan to exhaust the heat generated by the GPU. What I would like to see in BIOS is a fan control that I can make any of the fans dependent on GPU temperature.

[@henry3397]

Before watching your video… tons of others have already done this test. But it’s good to have someone with your exposure do it. The cliff notes: front mount is better, period. Front mount gives supreme cpu temps, while having a marginal effect on GPU temps. Top mount sacrifices CPU temps, for a tiny improvement to GPU temps. HOWEVER, when you have a blower style card, such as founders Nvidia, top mount can be a very good design choice, as this type of GPU will exhaust a significant amount of heat from the rear expansion port, and thus CPU temps won’t be as greatly affected.

[@dainiusvysniauskas2049]

Couple of counterarguments - front mount eats up a lot of space of the case, which can be an issue with modern graphics cards, also graphics cards are more temperature sensitive than CPUs. Generally, your CPU doesn't give a damn if it is running at 70c or 75c. GPUs tho? They care. I don't remember the exact number right now, but Nvidia GPUs start to limit their boost clocks at 65c or so. So the few degrees difference on GPU temp does matter more than few degrees difference on CPU.

[@Ntrglycrn]

Another test would be to have a pull/push fan configuration on your radiator. I have a 360 NZXT radiator in an Lian Li O11 Dynamic case, mounted in the vertical opening pushing air from the computer "room" towards to cable "room". In total I ruin 9 fans (12 if you would include the fans on the GPU) in my configuration: 3 120 mm push through the rad, 3 120 mm pull through the rad, 1 140 mm air intake on the bottom of the case, blowing towards the 7900 XT GPU and 2 on the top of the case, blowing air out. My AIO hosts do run upwards from the pump/cpu block towards the radiator, similair to the configuration of the computer next to Jay at 4:55.

[@richard-6920]

Front mount is cpu bias, so the gpu will get warmer air, top mount is gpu bias, so the cpu will get warmer air. Placement really depends on where you have thermal problems. Id expect a max 5c difference so its not super important. Chances are your intake /exhaust aren't the only intake /exhaust so the temp difference should be even less.

[@PeaJayTheGr8]

I am running my rad from the top as exhaust but did something a little unconventional. I have 2, 240mm fans as intake in the front, and I swapped my 120mm fan in the back to an intake instead of exhaust, hoping to create positive airflow instead of negative. Doing so, I've noticed that my CPU temps dropped about 4 degrees, so I am happy with this unconventional move. 👍

[@pandalife_gaming]

If we're assuming the rad is on an aio cooling the cpu, then my logic would be to have the rad as exhaust. As intake it would slightly heat up the air in the case but moreover it would impede airflow over other components that need fresh air. Cpu's (generally) use less wattage than gpu's and therefore would benefit less from additional cooling. Thus having cooler cpu temps matters less. But ultimately my guess would be that it doesn't make a major difference either way and you should just do what works for you. Edit aftet the video: Woot basic thermal dynamics lol. That said Jay's take at the end is kinda unnecessary. His own testing shows a ~4°C difference in temps for the gpu. Which roughly translates to a ~4°C difference in all other components as well. All for putting your rad somewhere else. That's reasonably conclusive that there is in fact a correct placement for your rad, assuming you have sufficient space. Also assuming you aren't running a 14900k and need that 2°C extra headroom lol

[@MrSecurity.]

It's simple. Put it where you want it. In my first custom build I had a 240 radiator. It could only fit in the front of my case so it's in the front and the hoses were at the top because I built it in 2017 and nobody was talking about air bubbles inside of radiators. Also, I thought it'd be a good place to put it cuz cold air would hit it from the front. My new build that I built last September. I put my 360 radiator on the top because I saw some of your videos talking about it and I also have the lian li 011 dynamic Evo and there's not a place to put it on the front. I could put it on the side but I didn't like the way that looked. I'm not like some expert PC maker but I would argue that it's really not that big a deal. I suggest people put it where they want it.

[@AlexraptorGameDev]

Biggest issue I've had with front mounted rads, is balancing airflow and noise, since a lot of the airspeed/energy is lost moving the air through the intake mesh and radiator. For my main system I'm exclusively using AIO cooled GPU's these days, with GPU rad mounted at the back and CPU rad mounted up top. That way all the internal components stay as cool as possible.

[@TurboLoveTrain]

The first rule of thermodynamics is don't talk about thermodynamics.

[@xCaptainApathyx]

I have a 360 in the front for intake, and a 360 in the top for exhaust, all fans in "push". Fan curve based on coolant temp sensor. Cooling a threadripper 1900x and a 1080ti just fine. The 40-50° air blowing through my case is not harming chipset, ssd, RAM, or anything else. at all.