Yeah I think it’s kind of dumb to test with a tempered glass side panel. I could have told you it would be worse and saved you the time spent making the video.
For my build in an NR200 I actually went with a NH-C14S because the case has side ventilation so it's pulling air straight from outside of the case. It's a downdraft cooler but it's a *massive* downdraft cooler. Works really well.
i know this is an ancient comment, but im doing a build in a NR200, and i can tell you, the case has clearance for up to 155mm cooler height, which means it could fit all the way up to a thermalright peerless assassin 120, which is one of the biggest and best coolers on the market. thereses absolutely no reason to go with a downdraft cooler, unless you just want to, and are ok with sacrificing potential cooling performance.
I liked the methodology, normalizing for noise and testing different fan configs. Only thing I would change is to use a more airflow focus case. The H510 relies heavily on negative pressure, favoring the GPU and "suffocating" the CPU, as shown by GN. Great job, nonetheless
I agree but in this situation it was more about normalizing things and making sure there's a controlled environment rather than focusing solely on airflow. But yeah I get it. Maybe another test with better ventilation.
AGREE. i actually appreciate testing inside a case but this case is "wost case scenario" (pun intended) if there is only a 3 degree temp diff in this shitty case with no airflow then it will probably be negligible difference in a good case.
@@HardwareCanucks I’m totally ok with the case selection, but WHY did you go one intake and one exhaust? This case works way better with two exhaust, and I’m wondering if a dual exhaust would work better with a downdraft style cooler.
Have to agree here, the case choice was not the best platform for this test. I get the idea of showing the worst case scenario, but that case is going to restrict the performance of any air cooler to a point where I don't think the numbers are useful for evaluation. Also, using a GPU that blows hot air up in front of the CPU is going to always favor a tower cooler because the air moved by the GPU is placed directly in front of the fan blades of a tower cooler, while it's parallel to the fans on a downdraft cooler.
Thank you so much for confirming the 180° Flip for AMD. As long as it doesn't block the Left Most DIMM Slot, I am happy. I would prefer the cooler on VRM instead of RAM. RAM is not as hot, and I have some decently tall RAM.
A case with a fan on the side for exhaust or intake would make a huge difference. In the tested case, the hot air is just bouncing off the glass. But most cases these days don't have a fan or mesh on the side so doesn't matter much.
Well I'm rarely this early to a video. Gotta say, this is a topic that seems very interesting to know more about. Edit: These were interesting results, I appreciate it.
This needs a part 2! You actually tested downdraft coolers with a glass sidepanel. I didn't actually think someone would do that because obviously it's going to be worse, since it can't get fresh air from a ventilated sidepanel.
I'd love to see these tests on lower-end components; motherboard with no VRM heatsink, 95W CPU heatload, ram with no heatsinks. That would be the real test. I think the tower cooler was better in these tests with a higher-end system because the tower cooler more effectively removes heat from the case, causing lower ambient temps and negating the benefit of the direct airflow the down-draft coolers provide. Thank you for the insight, HW Canucks team!
I opted for a Dark Rock TF in my NR200 build because it had biggest fans I could fit given my clearance, letting me run lower RPMs for less noise. So far I’m happy with it!
This is a fantastic video! For me, I have a NH-C14S modded to use LGA2011-3 Narrow brackets to fit on my ASRock X99E-itx with a 6950x. I have the fan mounted under the heat sink pushing up through the fins. I absolutely love this cooler and am now going to be researching that Dark Rock TF2, thank you!
I like downdraft cooler because it also cools VRM and NVMe SSD which usually placed right below thr CPU socket. The difference in SSD temp is quite noticeable as I see 20°C decrease in my Samsung SSD 960 EVO. As for VRM, it helps most basic motherboards which usually have smaller or even no heatsink. I like how good the original AMD Wraith cooler. Too bad AMD is just like Intel nowadays... they no longer provide stock cooling solution on high-end models. The midrange and lower end CPU also ended up with just basic cooling solution.
Personally, I like to have a top exhaust fan and two ingesting fans in the front. They don't have to be high speed, but fast enough to keep some air flow. One thing many people forget, if the back of the case is near a wall, there will be some heat backwash into the case. If you can get a cheap USB fan to clear that, it can help immensely.
neat. glad to see downdraft getting some attention! machines and more found some interesting interaction within the nr200 in which a downdraft helps compartmentalize and isolate temps between CPU and GPU. certainly case specific, though.
I am using a down draft cooler for aesthetic and component cooling reasons. I underestimated the CPU temperature rising, tbh, and I went from a budget tower cooler to a high end (!) down draft cooler. VRM/RAM and M.2 are cooler now, though. Edit: The main reason of this are rendering / calculation sessions where storage and memory are heavily used for a longer time.
This is a very well made video with excellent presentation and graphics, keep it up and i love how you used correct terminology of hypothesis and not theory!! You just earned a long subscription from me.
Thank you for posting. I have the Noctua C14S w/an R9 5900X on an ASUS ROG Crosshair VIII Impact mb (ITX). The cooler is pushing out air in a CM NR200 case. Plus an RTX 3080 Ti FE. My CPU temps stay about 67-70 degrees. GPU about 68-74 degrees. Very happy w/ those postings. Oh, I have 5 Noctua fans.
Love your videos usually, but this one feels essentially misleading or essentially useless: people looking at downdraft coolers usually will have vented side panels so they can utilize direct fresh air/exhaust. Further, the GPU is quite underpowered to really affect things. It would be much interesting to see how a 3070 or 3080 would fare. Keep up the good work!
Nice! I'm kinda old-school and prefer the look of downdraft coolers :D But I have to ask, could we see the results with flipped fans *and* a ventilated side panel? dumping that hot air directly outside should theoretically give even better results, at the cost of aesthetics...
Great test and thank you for the hard work :) I tested similar things with my current setup and can only imagine how long it took you with even more coolers. I actually use the DarkRock TF in my Cooler Master NR200 with mesh side panels (never tested it with glass tbh). I also tested tower cooler, until I finally settled for the TF. Setup is: 2 Fans on top - blowing out. 2 Fans on the bottom - blowing in (this is mainly for the GPU. I use the GPU with Morpheus Cooler and the bottom fans are the only fans for the GPU). DarkRock TF sucking fresh air right in from the side panel and bursting the air on the ITX board. In my scenario, the TF performed about equal for CPU and for VRM being the coolest. Theory: CPU cooler blowing on all components on the ITX board, while the fans on the top suck it right out. Since they are right next to it, the air has no chance to get caught somewhere. I guess in your test, it will always favour the tower cooler as the air flow is basically a straight line. In a not so small ITX case, like the NR200, the two top fans are so close to the cooler, that they heavily support this kind of cooler. But there might also be other reasons: heat load of the CPU; where the VRMs or other components are located, is there even a front panel with air intake or is it a case like the NR200, which only allows side ventilation, etc... Ofc, there might have been some flaws in my testing, as I just wanted to get it over with. Tower coolers were a Scythe and Alpenfohn; I only tested one of them turning the cooler 90 degrees and I never put the glass as a side panel on.
My primary point of disagreement is the case. The H510 is a poorly ventilated case and i believe that in a mesh fronted case the results will differ. It would be nice to have that tested. Regarding the GPU, in the H510 the graphics card draws air from the PCIe slot covers and not throug the front of the case. So the measurement of the GPU temperature is irrelevant beacuse the CPU cooler cannot disrupt the airflow coming to the GPU which might be an issue with a front ventilated case in the case of(pun intended) a positive pressure oriented chassis. Also the temperature deltas of tge fan positioning are so small that you can take them as insignificant. Feel free to change my mind
I actually own the Shadowrock TF2 cooler with the dual fan design. I chose it for aesthetic reasons mainly and also because I read reviews on it and watched many of them and it had comparable Temps to many tower coolers, actually besting some of them. In my experience, I've loaded this thing up with CPU intensive workloads, editing videos and gaming. Ryzen 3900X @ 4.45ghz, RTX 2070, 64GB RAM if system specs are needed for reference. Temps never got above 65°C for the CPU. Also, I used this in an OPEN AIR CHASSIS, like the Motiff Monument. So it's a beautiful build. I'd Def recommend the TF2 cooler, it's as good as any out there. Plus it looks AMAZING lol.
So, you need a Fan to the top of the case for a Downdraft cooler. and the temps will go down. just like how you've got a fan to the back of the case for a Tower cooler.
Cases with side ventilation used to be more common and you kind of want that for this case. The good slim ITX cases have ventilation directly over the CPU to aid the downdraft coolers
Did not include any front vented configurations, but the info that was collected & shared was VERY informative and eye-opening. I have a NR200 with a vented front panel (not glass) and a top-down Noctua NH-C14S cooler. The cooler has 2 fans pulling air top-down, and the other spot on the front panel that mount a fan bringing in fresh air. The top 2 fan are exhausting hot air. I also have a i7-10700k, Aorus 2060 Super with no case fans installed at the bottom. I am going to flip the blowing direction of all fans: the NH-C14S cooler fans will be bottom-up, front panel fan will exhaust and at the top fans will intake. Although heat rises, the force behind rising hot air is insignificant when compared to top fans intaking cool air. I will record temps in this second configuration and let you know what i find.
I personally prefer the downdraft coolers for the aesthetics and motherboard cooling, particularly if you have a case with a transparent side panel and because the horizontal fan orientation of downdraft coolers naturally fits in with the appearance of the rest of the PC. I've never liked the sight of a tower cooler when viewed from a transparent side panel. The AMD Wraith Prism cooler is especially beautiful to watch as it cycles through RGB colors and patterns.
the shadow rock tf is a weird one, all the gimmick of shadow rock line is being quieter than others hence the low density of fins in towers, though the tf version is so tight it looks like you have to spin the fan much faster to push through
@@HardwareCanucks it still affects acoustics negatively, of course it matters only when you care about it and this channel never was really concerned about it
Awesome timing. I was just starting to look at DownDraft Cooler to replace my Hyper 212 LED. Maybe look at the Shadow Rock LP as well. A much more budget friendly price point.
These style coolers have their fans set to pull air, so does that make them updraft? These can only be downdraft, if you make them so! These are top down or top flow coolers as the fans airflow orientation does not define them 😉
@Cum in belly button My research seems to yield different results. Based on BitWit Kyle's older video about the Shadow Rock LP, it performed as well as the Shadow Rock Slim. Which is a 160W TDP Cooler. Hardware Canucks found in the "The Best BUDGET CPU Coolers!" Video, that the lowly Pure Rock Slim 2 with a 130W TDP outperformed the Hyper 212s. Based on that, I would think the LP would be at least as good as the Hyper 212. And another issue with my Hyper 212 is: it has the very early AM4 mount, which only allows you to mount it Horizontally, with the airflow going up in a Pull Configuration. Which is not ideal. Of course. This video changed my mind for the most part, and I will probably go with a different Tower Cooler instead.
Thank you for this content. One of the most important things that you (and others) leave out of air cooler reviews is how they compare to stock coolers such as AMD's Wraith series and Intel equivalents. I know it is time consuming to test coolers, but your findings are of little value to us if you don't tell us how stock coolers perform on your test system so we have a reasonable comparison. The ideal solution would be to perform all tests in a case rather than an open test bench as this is more representative of what we would have. Obviously that's impractical and there is no such thing as a 'typical' PC case. So I think the least you could do is to make sure to ALWAYS test stock coolers on the same test bench so we have something to compare your results to.
Not surprising that downdraft coolers work worse in cases designed for tower coolers. Would make sense to test in a case that can be configured with a solid or vented side panel like the NR200. Then you could basically test a tower cooling in an ideal config vs a downdraft cooler in an ideal config in the same case.
In short, only get downdraft coolers if you're space limited. Might be worth your while too if you cut an opening in the case so that it can vent the hot air straight out.
I am using a downdraft cooler the now discontinued nh-c14. The thing with after market downdraft coolers that you have to take into account is that it used for niche cases as you have to take into account the airflow of your case. Front intake, rear and top exhaust with either a mesh side panel or a side panel with intake fans. Unlike most tower cooler that are mostly plug and play on most modern cases as they were designed with tower coolers in mind ( they are the most common aftermarket coolers ).
You mentioned the Dark Rock TF2 is around the same height as a 92mm cooler so actually that is a comparison I would like to see. What happens when you have say space that prevents a 120mm tower cooler (so like sub 145mm)...is a 92mm tower better or the premium down-draft cooler? And like others mentioned I would love to see a comparison with either just a more air-flow focused case and/or side-mesh case. Heck maybe even a video on optimizing airflow for a down-draft cooler as opposed to a tower cooler.
I'd definitely be interested in a test of the coolers in different case styles. I would hypothesize that you might see a larger difference for a tower cooler with a mesh front case with a few more front fans, and as was mentioned would be interesting to see if a mesh side panel helped the downdraft cooler.
The only option to effectively deal with multidirectional flows so that the cold does not collide with the hot one is to remove the hot flow through the diverted corrugated channels (for example, like kitchen hoods)! And make sure that no NEGATIVE PRESSURE is created inside the case!
7:16 because side fans, as intake fans, blowing directly onto the motherboard and other components is superior to front intake fans. If you blow air directly onto a downdraft cooler, the air is fed directly into the cpu cooler.
I wonder if the downdraft cooler was rotated (if possible) in such way where case fan/s could push trough the radiator fins instead of the blocked top of radiator. I don't know maybe that is done on purpose so air wouldn't collide within radiator but I am curious about that. I feel the test set up is optimal for tower coolers. I wonder if there was another test setup configuration that would be more thought out for downdraft cooler in which it would have better results... like adding another exhaust fan to pull out hot air from VRM area, maybe it needs additional case fan for more fresh air circulation within case to unlock full potential of downdraft cooler. And if there are better test results in 2nd test setup it would be nice if there was comparison of each cooler in both test setups and comparing the results. I feel that would be really helpful. ^^
Thanks. I am considering a downdraft cooler. Reason the size of a tower cooler and the weight of it. Being physically handicapped I cannot hold a big heavy cooler in place with one hand whilst struggling to secure the screws with a long fiddly screwdriver. I was hoping that a downdraft cooler would be A lighter, and B require a shorter screwdriver to attach it, which would be easier for me to handle. Peerhaps you could show the differences and consider that not everyone has the strength to hold large chunks of cooler in mid air.
Great testing. I actually think that downdrafts are mainly for space constraints nowadays. However, there are still people who do not use glass side panels and prefer the ones with the side mesh. Guess a good follow-up will be trying different style of cases?
Thanks for all the testing, but I think the video falls very short for viewers without a succinct summary at the end, and there were no clear charts for comparison synopsis. Maybe I’m not a good video consumer, but I want to see all the data at once to use for my planned use case.
Should have put an exhaust fan at the top, especially for the top-down cooler blowing towards the MOBO. An extra exhaust fan right there would move that hot air right out and probably give the best temps all around.
I have a CPU Tower Cooler. I bought one ID IS-55 because I don't like the look of the tower. It gets in the way of accessing other components. But I'm switching primarily for aesthetic reasons. Thanks to the channel analysis I opted for the IS-55.
I vertical mounted my gpa in my phanteks 719, which exhausted air towards my CPU socket. Went with the dark rock TF, and packed that case with 14 fans. There’s so much airflow with those fans running at 650rpm that even under heavy load everything stays cool
Love these videos on air coolers but I have to object to using that case for the comparison. The tower cooler has a direct line of sight with the front and rear fans for maximum airflow. The top down coolers are dealing with glass which obstructs air in both configurations. I'm running a Coolermaster NR200 case with a Noctua NH-L12S cooler with a side fan pushing air into the heatsink and 2 case fans exhausting air out the top of the case and the temps are superb. You can use a Fuma 2 in this case which is a great heatsink and would make for a better comparison.
I do understand but this was more about achieving a controlled environment in order to normalize results rather than "stacking the deck" one way or another. If anything it shows more of a "worst case" scenario wherein none of the coolers had a specific advantage.
@@HardwareCanucks I would have to disagree with the idea that none of the coolers had a specific advantage because the tower cooler is having fresh air pushed to it by the front case fans while the rear fan pulls the heated air out of the case. The top down cooler is pulling in air with a piece of glass blocking outside air and the gpu is heating nearby air which gets pulled into the heat sink fan. At he very least the glass panel should be removed and the test run again to see what differences, if any, occur.
A downdraft cooler was the solution for my sleeper PC. My old In Win V-508 series case was too narrow for a 120mm tower despite being mATX. I tested with a Cooler Master Hyper 212 and couldn't close the case. I also tested with the Be Quiet Dark Rock TF2 but I could only mount it one way on my particular motherboard due to the VRM heatsink layout and it physically rested against ram in the closest slot to the CPU, the heat pipes pushed against the GPU backplate while the bulk of the cooler blocked cables from the top-mounted PSU and the 5.25 bays. However, my case came with a side duct, which after a little adjustment sat right above the CPU cooler. Even so, I quickly replaced the stock Intel cooler with a Noctua NH-9x65. I am using a hot CPU too (17-11700F). I replaced the stock 3-pin 92mm exhaust fan in my case with a 92mm pwm exhaust fan, added a single 80mm pwm intake fan and modified one of those 3x40mm 'hard drive' 5.25" bay coolers with upgraded 40x28mm pwm fans and a Noctua low noise adapter because otherwise it's intolerable; and a 92mm exhaust fan. CPU temps are 40 C idle and have sometimes broke 80 when abusing the thing.
Good video, but thought it was a bit odd to benchmark the downdraft without ventilated side panels. It clearly doesn't get access to fresh air, and in the exhaust orientation the hot air doesn't escape.
I've been using a Scythe Grand Kama Cross 2 inside a NZXT Phantom 530 since 2013. The main reasons were the low rpm 14mm fan at the top, which is really silent and does the job just fine (i5-4670k) and the RAM clearance. Oh, and it looks cool!
11:01 that moment when you know there's others having in mind the same question and the prove :) thank you So I will share idea for that housing - Cooler Master HAF XB EVO. Very old, but perfect for DARK ROCK TF 2 and easy to rebuild front panel. Cool case for DIY project with TF 2
I use a Dark Rock TF updraft in a CM HAF XB box case. Dark Rock TF has 2 fans. Updraft setting is still push-pull. And lower fan is low enough to suck hot air from motherboard. HAF XB has a big mesh just above CPU and GPU. Both CPU and GPU heat will directly go up, out of case. I use DR TF on a 5900X, good for gaming and light video editing. IMO, downdraft coolers are for box or horizontal cases, sucking cool air from above or push hot air upward. They are not for tower cases.
I specifically chose a downdraft cooler for a retro build that had a fan on the side panel, which reduced clearance and complimented the cooler well, but for a modern full-sized system like my daily driver, I’ll choose a tower cooler. I'll get whatever is the right tool for the job.
Great video Mike! Although, what would the downdraft coolers look like if there was ventilation from the side panel? I would think that'd have a huge advantage to them over the tower coolers. Mind you, most cases have gone with solid side panels now, so might be hard to find newer ones that do hehe I noticed you had images of an open test bench, does that mean you have another video coming out to see which ones just cools better overall?
Great question. It really depends on how this content does TBH. The main problem we have is a real lack of availability of cases that support higher tower coolers while also having side ventilation.
I would suggest to use NR200 with vent side panel. It has decent CPU cooler clearance. I remember watching few other channels tested both and tower coolers still performed better at every test.
I agree that the downdraft coolers were greatly handicapped by the solid side panel. I think it would be more relevant if tests were also conducted on an open bench to measure the real potential of each cooler and then tested in the specific case for comparison.
i've noticed that the direction of the fin array when installed can have an effect on airflow. having them positioned horizontally instead of vertically seems to help exhaust air out of the side instead of down towards the hot gpu where it will clash with the air being pushed up from there and possibly create a lot of turbulence in your case.
The test setup is skewed against downdraft coolers, you need to move the case exhaust fan to the top (perpendicular to heatsink fins) for downdraft coolers to be a little more efficient. Some heat is being recirculated in the case. Best would be to have a side intake fan if possible. Anyway, downdraft coolers have a good use for vented side panels like NR200, NCase M1, Cougar QBX, etc. They perform better with fresh intake directly going to the heatsink & components usually get 5-10°C lower temperatures. Just make sure you have an exhaust fan so heat doesn't get recirculated inside the case.
Not sure if your still looking at comments or not. Looking back at this video I see a flaw in the testing that you touched on, down draft coolers are not designed to work in modern cases due to the direction of air flow. A more "accurate" test for direct cooler comparisons would be to use both on an open test bench and see how they directly effect temps. This would remove the "advantage" you gave the tower coolers with the use of a case optimized for their air flow pattern.
I have the Dark Rock TF with two fans exhausting air out the top of the case which is also the way the fins on the CPU cooler are aligned. I think this would make a noticeable difference in temperatures compared to the setup in this test. Having a fan at the back of the case favours the tower style cooler because that is the way the fins are aligned but the fan at the back of the case would do very little for a downdraft cooler if the fins are aligned to the top. The Dark Rock TF is complete overkill for my system which is Be Quiet themed and I mostly got if for the looks but I think having fans at the top of the case is the way to go with this kind off CPU cooler if the fins are aligned that way.
High-end motherboard isn't really the best choice to have conclusive results on VRM temperatures. It would be much better to have such a test done on a mid-tier or garbage motherboard with a VRM that heats up to at least 80 degrees - that's where helping to cool the VRM starts to make a difference.
I can add my 2 cents into this conversation about both coolers, the dark rock 4 and dark rock tf2 as I had both installed (currently sticking to TF2). I have a Mini cube case with a single 140mm fan on the front, 120mm on the back and a chance to add a 120mm top fan as well. At the front I have Noctua Industrial 3000pwm set to 30% under 80 C and it goes up from there as required (I don't mind the fan noise, I actually like the swoosh/white noise at 30-45% to have in background when I'm working). On the back I have Redux 1700pwm set to silent. My motherboard lies flat in the case therefore with the TF2 I can access top airflow. My specs are Ryzen 7 5800X3D, B555M steel legend. Both coolers were configured manually to best performance. When I was running Dark Rock 4 my CPU temps at 105W was around 81-82 degrees under full load stress test after 30 minutes. VRM temp was around 60-61 C and ram 39-41 C. When I switched to TF2 CPU temps at 105W on my CPU are now topping a 78 (highest one I've noticed under the same stress test for the same amount of time), VRM temps dropped to 48(!!!) C and Ram temperature went down to 34C. I have actually noticed a drop in GPU (5700XT) temp by 3 degrees when using TF2. It really depends a lot on the case and how you can utilise airflow, and in case of TF2 when I added a extra top fan (noctua redux 1700) the temps dropped by about 2 degrees more but to keep it long term its a bit bad due to clearance issue, Soon I'm actually going to be upgrading the GPU to 7900XTX which I have tested before on both coolers and under heavy gaming in the same conditions I have also noticed nearly 5 degrees difference when using TF2 instead of Dark Rock 4. In my conclusion, if you do have a case that can utilise airflow from the top, it can really be beneficial as the air will actually cool your VRM and Ram and potentially GPU as well depending on your configuration. Overall, TF2 is an excellent cooler but I would say its more oriented for cases where the fans can get direct access to airflow.
Anyone remember those cases from years ago with side intakes that had actual plastic funnels leading to the CPU? Was probably back in the days when downdraft coolers (like the Zalman 'flower') were commonplace. Feels like cases today could learn lessons from those days, even if we weren't running 190 Watt CPUs (much) back then :)
yeah. 80mm /120mm fan on the acrylic side. those can cool the Athlon XP's and P4's back then. They don't look great but certainly helped a lot. Plus the PSU acts as vents for these coolers as it is located at the top.
With so few cases these days sporting mesh side panels, not sure what an "optimal" case would look like that would support all these coolers for normalized testing.
In addition to a classic mid tower air flow I have a big 500rpm 1700mm Noctua fan inside the tower door. So it sucks in additional cold air slowly on the whole board from basically up the mainboard, while there is a constant "classic" flow from the front to back. I could imagine in my case a down draft cooler might make a lot of sense and will be efficient. What I dont like about dense fins is that it gets easier blocked by dust over time though. If u dont clean it, over the years there will be a layer of dust almost completely blocking all the air flow towards the CPU :) Oh and I heard m.2 SSDs work better when under some heat..So that might be a + for downdraft coolers too..
The higher temperatures on the VRM and memory you've measured with downdraft coolers is probably due to 2 factors: hot air from the CPU being blasted over them (as you mentioned) but also the fact that downdraft coolers disrupt the "normal" airflow in a case. By the looks of it, you used 2 intake fans at the front and 1 exhaust on the back, pretty standard and it gives you a front to back airflow. The downdraft cooler gives a lot of turbulent air in that flow, not helping cooling. I generally would only use downdraft coolers in 2 situations 1) Home theater pc's which you want to keep as quiet as possible, thus eliminating as many fans as can be. When using an APU, the CPU cooler alone should be enough for the entire system (the rest being passive exhaust) and in that case some extra airflow over the VRM can help, certainly on lower tier motherboards who sometimes lack VRM heatsinks 2) Office pc's where, just as with the home theater pc, you don't want extra fans in the case. A third option would be if you want a specific downdraft cooler for looks, but then you have to except the drawbacks.
I constantly had to stare at that empty exhaust fan spot at the top 🤔 Backside GPU Temp differences, expectations RTX3090 rear mem temps...ah nope. Well done vid tho 👍
I had a Scythe tower cooler and I switched it to AMD prism cooler just to get VRM temps down. I have a Ryzen 5600 on a A320 motherboard so yeah. VRM temps measured from HWinfo dropped from 110 C to 75C! Test was Cinebench R23.
I have been using the Original Dark Rock TF since 2016 and I love it both in looks and performance and I've been waiting patiently to see a test like this. I only wish it had a Am4 bracket.
If only there existed some kind of visual tool for summarizing the numerical results in one image to allow for comparisons to be made more efficiently.
Maybe for a part two test in a case with a mesh side panel? Would be interesting to see how much better the downdraft coolers perform there.
Definitely. SFX with breathable panel or is where the downdraft coolers should shine.
If there's enough demand...you KNOW I'll do it. 🙃 But if not, onto the next thing.
@@HardwareCanucks I would be interested in a sequel to this video. A case with a front mesh panel would be good to include.
Yeah I think it’s kind of dumb to test with a tempered glass side panel. I could have told you it would be worse and saved you the time spent making the video.
I think you could do a better test with the NR200 to really see the best use case, and it has good cooler compatibility.
ive been helping a friend budget his pc build and this channel has really helped me find ways to cut costs without cutting corners
Good to know. Thanks for the feedback!
For my build in an NR200 I actually went with a NH-C14S because the case has side ventilation so it's pulling air straight from outside of the case. It's a downdraft cooler but it's a *massive* downdraft cooler. Works really well.
And it's gonna be even better if you mount 140s in the side.
i know this is an ancient comment, but im doing a build in a NR200, and i can tell you, the case has clearance for up to 155mm cooler height, which means it could fit all the way up to a thermalright peerless assassin 120, which is one of the biggest and best coolers on the market. thereses absolutely no reason to go with a downdraft cooler, unless you just want to, and are ok with sacrificing potential cooling performance.
One of the best hardware related videos I've watched this month. Thanks a lot Mike, definitely worth it for you to wake up that early!
Thanks!
Would love to see these compared in a case with side panel intakes. There's a reason top flow was more common in the days before windowed side panels
I liked the methodology, normalizing for noise and testing different fan configs. Only thing I would change is to use a more airflow focus case. The H510 relies heavily on negative pressure, favoring the GPU and "suffocating" the CPU, as shown by GN. Great job, nonetheless
I agree but in this situation it was more about normalizing things and making sure there's a controlled environment rather than focusing solely on airflow. But yeah I get it. Maybe another test with better ventilation.
AGREE. i actually appreciate testing inside a case but this case is "wost case scenario" (pun intended) if there is only a 3 degree temp diff in this shitty case with no airflow then it will probably be negligible difference in a good case.
@@HardwareCanucks I’m totally ok with the case selection, but WHY did you go one intake and one exhaust? This case works way better with two exhaust, and I’m wondering if a dual exhaust would work better with a downdraft style cooler.
@@HardwareCanucks if you do this please record GPU temps as well to see the difference if there is any.
Thank you for the content!
Have to agree here, the case choice was not the best platform for this test. I get the idea of showing the worst case scenario, but that case is going to restrict the performance of any air cooler to a point where I don't think the numbers are useful for evaluation.
Also, using a GPU that blows hot air up in front of the CPU is going to always favor a tower cooler because the air moved by the GPU is placed directly in front of the fan blades of a tower cooler, while it's parallel to the fans on a downdraft cooler.
Thank you so much for confirming the 180° Flip for AMD.
As long as it doesn't block the Left Most DIMM Slot, I am happy.
I would prefer the cooler on VRM instead of RAM. RAM is not as hot, and I have some decently tall RAM.
Yeah, this is a big deal that a lot of folks miss. If there's VRM heatsink clearance issue, just rotate.
A case with a fan on the side for exhaust or intake would make a huge difference. In the tested case, the hot air is just bouncing off the glass. But most cases these days don't have a fan or mesh on the side so doesn't matter much.
Well I'm rarely this early to a video. Gotta say, this is a topic that seems very interesting to know more about.
Edit: These were interesting results, I appreciate it.
This needs a part 2! You actually tested downdraft coolers with a glass sidepanel. I didn't actually think someone would do that because obviously it's going to be worse, since it can't get fresh air from a ventilated sidepanel.
I'd love to see these tests on lower-end components; motherboard with no VRM heatsink, 95W CPU heatload, ram with no heatsinks. That would be the real test.
I think the tower cooler was better in these tests with a higher-end system because the tower cooler more effectively removes heat from the case, causing lower ambient temps and negating the benefit of the direct airflow the down-draft coolers provide. Thank you for the insight, HW Canucks team!
With much respect, I thank you for all efforts you put for this quality content!
I opted for a Dark Rock TF in my NR200 build because it had biggest fans I could fit given my clearance, letting me run lower RPMs for less noise. So far I’m happy with it!
This is a fantastic video! For me, I have a NH-C14S modded to use LGA2011-3 Narrow brackets to fit on my ASRock X99E-itx with a 6950x. I have the fan mounted under the heat sink pushing up through the fins. I absolutely love this cooler and am now going to be researching that Dark Rock TF2, thank you!
I like downdraft cooler because it also cools VRM and NVMe SSD which usually placed right below thr CPU socket. The difference in SSD temp is quite noticeable as I see 20°C decrease in my Samsung SSD 960 EVO. As for VRM, it helps most basic motherboards which usually have smaller or even no heatsink.
I like how good the original AMD Wraith cooler. Too bad AMD is just like Intel nowadays... they no longer provide stock cooling solution on high-end models. The midrange and lower end CPU also ended up with just basic cooling solution.
Personally, I like to have a top exhaust fan and two ingesting fans in the front. They don't have to be high speed, but fast enough to keep some air flow. One thing many people forget, if the back of the case is near a wall, there will be some heat backwash into the case. If you can get a cheap USB fan to clear that, it can help immensely.
This is a really unique content , great work Mike ! Also loved the explanation and apresentation of the temps etc thank you !
Thanks!
neat. glad to see downdraft getting some attention!
machines and more found some interesting interaction within the nr200 in which a downdraft helps compartmentalize and isolate temps between CPU and GPU. certainly case specific, though.
I am using a down draft cooler for aesthetic and component cooling reasons. I underestimated the CPU temperature rising, tbh, and I went from a budget tower cooler to a high end (!) down draft cooler. VRM/RAM and M.2 are cooler now, though.
Edit: The main reason of this are rendering / calculation sessions where storage and memory are heavily used for a longer time.
This is a very well made video with excellent presentation and graphics, keep it up and i love how you used correct terminology of hypothesis and not theory!! You just earned a long subscription from me.
Thank you for posting. I have the Noctua C14S w/an R9 5900X on an ASUS ROG Crosshair VIII Impact mb (ITX). The cooler is pushing out air in a CM NR200 case. Plus an RTX 3080 Ti FE. My CPU temps stay about 67-70 degrees. GPU about 68-74 degrees. Very happy w/ those postings. Oh, I have 5 Noctua fans.
Love your videos usually, but this one feels essentially misleading or essentially useless: people looking at downdraft coolers usually will have vented side panels so they can utilize direct fresh air/exhaust.
Further, the GPU is quite underpowered to really affect things. It would be much interesting to see how a 3070 or 3080 would fare.
Keep up the good work!
Nice! I'm kinda old-school and prefer the look of downdraft coolers :D But I have to ask, could we see the results with flipped fans *and* a ventilated side panel? dumping that hot air directly outside should theoretically give even better results, at the cost of aesthetics...
Great test and thank you for the hard work :)
I tested similar things with my current setup and can only imagine how long it took you with even more coolers.
I actually use the DarkRock TF in my Cooler Master NR200 with mesh side panels (never tested it with glass tbh). I also tested tower cooler, until I finally settled for the TF.
Setup is: 2 Fans on top - blowing out. 2 Fans on the bottom - blowing in (this is mainly for the GPU. I use the GPU with Morpheus Cooler and the bottom fans are the only fans for the GPU). DarkRock TF sucking fresh air right in from the side panel and bursting the air on the ITX board.
In my scenario, the TF performed about equal for CPU and for VRM being the coolest. Theory: CPU cooler blowing on all components on the ITX board, while the fans on the top suck it right out. Since they are right next to it, the air has no chance to get caught somewhere.
I guess in your test, it will always favour the tower cooler as the air flow is basically a straight line. In a not so small ITX case, like the NR200, the two top fans are so close to the cooler, that they heavily support this kind of cooler. But there might also be other reasons: heat load of the CPU; where the VRMs or other components are located, is there even a front panel with air intake or is it a case like the NR200, which only allows side ventilation, etc...
Ofc, there might have been some flaws in my testing, as I just wanted to get it over with. Tower coolers were a Scythe and Alpenfohn; I only tested one of them turning the cooler 90 degrees and I never put the glass as a side panel on.
Sir, you are the best presenting all of these air coolers, you work is greatly appreciated.thanks!
My primary point of disagreement is the case. The H510 is a poorly ventilated case and i believe that in a mesh fronted case the results will differ. It would be nice to have that tested. Regarding the GPU, in the H510 the graphics card draws air from the PCIe slot covers and not throug the front of the case. So the measurement of the GPU temperature is irrelevant beacuse the CPU cooler cannot disrupt the airflow coming to the GPU which might be an issue with a front ventilated case in the case of(pun intended) a positive pressure oriented chassis. Also the temperature deltas of tge fan positioning are so small that you can take them as insignificant. Feel free to change my mind
This is a really interesting and informative piece. Thanks for that and sorry you had to get up so early to make it.
I'm an early bird!
I actually own the Shadowrock TF2 cooler with the dual fan design. I chose it for aesthetic reasons mainly and also because I read reviews on it and watched many of them and it had comparable Temps to many tower coolers, actually besting some of them. In my experience, I've loaded this thing up with CPU intensive workloads, editing videos and gaming. Ryzen 3900X @ 4.45ghz, RTX 2070, 64GB RAM if system specs are needed for reference. Temps never got above 65°C for the CPU. Also, I used this in an OPEN AIR CHASSIS, like the Motiff Monument. So it's a beautiful build. I'd Def recommend the TF2 cooler, it's as good as any out there. Plus it looks AMAZING lol.
you guys explain it wayyy better than be quite itself. Good job guys !
So, you need a Fan to the top of the case for a Downdraft cooler. and the temps will go down. just like how you've got a fan to the back of the case for a Tower cooler.
So that's where that poll went.
You got it!
Cases with side ventilation used to be more common and you kind of want that for this case. The good slim ITX cases have ventilation directly over the CPU to aid the downdraft coolers
Earlier-model XB1 is the same way, although with a top exhaust instead of a top intake, when a top intake would work better for it.
The case seems to be designed for tower cooler airflow, so it is obvious that downdraft cooler would perform worse.
Did not include any front vented configurations, but the info that was collected & shared was VERY informative and eye-opening. I have a NR200 with a vented front panel (not glass) and a top-down Noctua NH-C14S cooler. The cooler has 2 fans pulling air top-down, and the other spot on the front panel that mount a fan bringing in fresh air. The top 2 fan are exhausting hot air. I also have a i7-10700k, Aorus 2060 Super with no case fans installed at the bottom.
I am going to flip the blowing direction of all fans: the NH-C14S cooler fans will be bottom-up, front panel fan will exhaust and at the top fans will intake. Although heat rises, the force behind rising hot air is insignificant when compared to top fans intaking cool air. I will record temps in this second configuration and let you know what i find.
The results are in... CPU Max during gameplay / VRM Max / RAM Max / GPU Max.
Original Set-Up: 57 / 52 / 49 / 70
Updated Set-Up: 59 / 44 / 42 / 72
I personally prefer the downdraft coolers for the aesthetics and motherboard cooling, particularly if you have a case with a transparent side panel and because the horizontal fan orientation of downdraft coolers naturally fits in with the appearance of the rest of the PC. I've never liked the sight of a tower cooler when viewed from a transparent side panel.
The AMD Wraith Prism cooler is especially beautiful to watch as it cycles through RGB colors and patterns.
I loved the testing done here, really shows the difference based on the type of cooler and configuration. Awesome video!
This guy have already answered all of my questions!
Thanks. Great test and very informative.
A repeat with a mesh would be interesting, but what you tested covered probably most people's setup.
the shadow rock tf is a weird one, all the gimmick of shadow rock line is being quieter than others hence the low density of fins in towers, though the tf version is so tight it looks like you have to spin the fan much faster to push through
Actually that's one of the reasons the put a larger, higher relative static pressure fan on the TF series. It's compensation for the denser fin array.
@@HardwareCanucks it still affects acoustics negatively, of course it matters only when you care about it and this channel never was really concerned about it
Awesome timing. I was just starting to look at DownDraft Cooler to replace my Hyper 212 LED.
Maybe look at the Shadow Rock LP as well. A much more budget friendly price point.
These style coolers have their fans set to pull air, so does that make them updraft? These can only be downdraft, if you make them so! These are top down or top flow coolers as the fans airflow orientation does not define them 😉
@Cum in belly button My research seems to yield different results.
Based on BitWit Kyle's older video about the Shadow Rock LP, it performed as well as the Shadow Rock Slim. Which is a 160W TDP Cooler.
Hardware Canucks found in the "The Best BUDGET CPU Coolers!" Video, that the lowly Pure Rock Slim 2 with a 130W TDP outperformed the Hyper 212s.
Based on that, I would think the LP would be at least as good as the Hyper 212.
And another issue with my Hyper 212 is: it has the very early AM4 mount, which only allows you to mount it Horizontally, with the airflow going up in a Pull Configuration. Which is not ideal.
Of course. This video changed my mind for the most part, and I will probably go with a different Tower Cooler instead.
I have a Shadow Rock LP to cool my 125W FX-6350 and it does a great job. I have to cool the VRMs because it's the maximum my mobo can take.
Wow I always thought down draft coolers would be better overall. I learned something today. Thank you for the video!
Thank you for this content. One of the most important things that you (and others) leave out of air cooler reviews is how they compare to stock coolers such as AMD's Wraith series and Intel equivalents. I know it is time consuming to test coolers, but your findings are of little value to us if you don't tell us how stock coolers perform on your test system so we have a reasonable comparison. The ideal solution would be to perform all tests in a case rather than an open test bench as this is more representative of what we would have. Obviously that's impractical and there is no such thing as a 'typical' PC case. So I think the least you could do is to make sure to ALWAYS test stock coolers on the same test bench so we have something to compare your results to.
Actually check out the Budget Cooler Roundup. We did test with the stock AMD coolers as well. :)
Not surprising that downdraft coolers work worse in cases designed for tower coolers. Would make sense to test in a case that can be configured with a solid or vented side panel like the NR200. Then you could basically test a tower cooling in an ideal config vs a downdraft cooler in an ideal config in the same case.
I never use a downdraft cooler in any of my customer builds unless their chassis has an opening in the side panel for direct cold air intake.
Thanks for breaking down these coolers
In short, only get downdraft coolers if you're space limited. Might be worth your while too if you cut an opening in the case so that it can vent the hot air straight out.
I am using a downdraft cooler the now discontinued nh-c14. The thing with after market downdraft coolers that you have to take into account is that it used for niche cases as you have to take into account the airflow of your case.
Front intake, rear and top exhaust with either a mesh side panel or a side panel with intake fans. Unlike most tower cooler that are mostly plug and play on most modern cases as they were designed with tower coolers in mind ( they are the most common aftermarket coolers ).
I feel as though the video could have just started at 2:32. Thank you for covering this subject though.
You mentioned the Dark Rock TF2 is around the same height as a 92mm cooler so actually that is a comparison I would like to see. What happens when you have say space that prevents a 120mm tower cooler (so like sub 145mm)...is a 92mm tower better or the premium down-draft cooler?
And like others mentioned I would love to see a comparison with either just a more air-flow focused case and/or side-mesh case.
Heck maybe even a video on optimizing airflow for a down-draft cooler as opposed to a tower cooler.
I'd definitely be interested in a test of the coolers in different case styles. I would hypothesize that you might see a larger difference for a tower cooler with a mesh front case with a few more front fans, and as was mentioned would be interesting to see if a mesh side panel helped the downdraft cooler.
Love your videos, Mike!
Amazing video! Very clear and concise and very useful for people needing info about this specific topic. Loving it! Keep it up 👍
Thank you!! This is exactly what I was looking for! 😭
Thanks a lot for this, now it depends on the situation for each PC and I can apply all of these info
The only option to effectively deal with multidirectional flows so that the cold does not collide with the hot one is to remove the hot flow through the diverted corrugated channels (for example, like kitchen hoods)! And make sure that no NEGATIVE PRESSURE is created inside the case!
7:16 because side fans, as intake fans, blowing directly onto the motherboard and other components is superior to front intake fans. If you blow air directly onto a downdraft cooler, the air is fed directly into the cpu cooler.
i love my Noctua NH-L9a. it open so much room in my case and enables the full case airflow to work.
I wonder if the downdraft cooler was rotated (if possible) in such way where case fan/s could push trough the radiator fins instead of the blocked top of radiator. I don't know maybe that is done on purpose so air wouldn't collide within radiator but I am curious about that.
I feel the test set up is optimal for tower coolers. I wonder if there was another test setup configuration that would be more thought out for downdraft cooler in which it would have better results... like adding another exhaust fan to pull out hot air from VRM area, maybe it needs additional case fan for more fresh air circulation within case to unlock full potential of downdraft cooler. And if there are better test results in 2nd test setup it would be nice if there was comparison of each cooler in both test setups and comparing the results. I feel that would be really helpful. ^^
Thanks. I am considering a downdraft cooler. Reason the size of a tower cooler and the weight of it. Being physically handicapped I cannot hold a big heavy cooler in place with one hand whilst struggling to secure the screws with a long fiddly screwdriver.
I was hoping that a downdraft cooler would be A lighter, and B require a shorter screwdriver to attach it, which would be easier for me to handle.
Peerhaps you could show the differences and consider that not everyone has the strength to hold large chunks of cooler in mid air.
Great testing. I actually think that downdrafts are mainly for space constraints nowadays. However, there are still people who do not use glass side panels and prefer the ones with the side mesh. Guess a good follow-up will be trying different style of cases?
Thanks for all the testing, but I think the video falls very short for viewers without a succinct summary at the end, and there were no clear charts for comparison synopsis. Maybe I’m not a good video consumer, but I want to see all the data at once to use for my planned use case.
Should have put an exhaust fan at the top, especially for the top-down cooler blowing towards the MOBO. An extra exhaust fan right there would move that hot air right out and probably give the best temps all around.
I have a CPU Tower Cooler. I bought one ID IS-55 because I don't like the look of the tower. It gets in the way of accessing other components. But I'm switching primarily for aesthetic reasons. Thanks to the channel analysis I opted for the IS-55.
I vertical mounted my gpa in my phanteks 719, which exhausted air towards my CPU socket. Went with the dark rock TF, and packed that case with 14 fans. There’s so much airflow with those fans running at 650rpm that even under heavy load everything stays cool
thats smells like something that needs to be tested in an haf x evo.
Love these videos on air coolers but I have to object to using that case for the comparison. The tower cooler has a direct line of sight with the front and rear fans for maximum airflow. The top down coolers are dealing with glass which obstructs air in both configurations. I'm running a Coolermaster NR200 case with a Noctua NH-L12S cooler with a side fan pushing air into the heatsink and 2 case fans exhausting air out the top of the case and the temps are superb. You can use a Fuma 2 in this case which is a great heatsink and would make for a better comparison.
I do understand but this was more about achieving a controlled environment in order to normalize results rather than "stacking the deck" one way or another. If anything it shows more of a "worst case" scenario wherein none of the coolers had a specific advantage.
@@HardwareCanucks I would have to disagree with the idea that none of the coolers had a specific advantage because the tower cooler is having fresh air pushed to it by the front case fans while the rear fan pulls the heated air out of the case. The top down cooler is pulling in air with a piece of glass blocking outside air and the gpu is heating nearby air which gets pulled into the heat sink fan. At he very least the glass panel should be removed and the test run again to see what differences, if any, occur.
A downdraft cooler was the solution for my sleeper PC. My old In Win V-508 series case was too narrow for a 120mm tower despite being mATX. I tested with a Cooler Master Hyper 212 and couldn't close the case. I also tested with the Be Quiet Dark Rock TF2 but I could only mount it one way on my particular motherboard due to the VRM heatsink layout and it physically rested against ram in the closest slot to the CPU, the heat pipes pushed against the GPU backplate while the bulk of the cooler blocked cables from the top-mounted PSU and the 5.25 bays. However, my case came with a side duct, which after a little adjustment sat right above the CPU cooler. Even so, I quickly replaced the stock Intel cooler with a Noctua NH-9x65. I am using a hot CPU too (17-11700F). I replaced the stock 3-pin 92mm exhaust fan in my case with a 92mm pwm exhaust fan, added a single 80mm pwm intake fan and modified one of those 3x40mm 'hard drive' 5.25" bay coolers with upgraded 40x28mm pwm fans and a Noctua low noise adapter because otherwise it's intolerable; and a 92mm exhaust fan. CPU temps are 40 C idle and have sometimes broke 80 when abusing the thing.
Good video, but thought it was a bit odd to benchmark the downdraft without ventilated side panels. It clearly doesn't get access to fresh air, and in the exhaust orientation the hot air doesn't escape.
Saw a review on a cheapo tower cooler, the fan was a bit over sized and mounted very low allowing for decent gem cooling.
I like the look of a downdraft cooler without the fan shown... just plain industrial fin stack with a fan underneath
You need your comparisons listed next to each other in the same frame. Going back and forth to see what changed is a HUGE pain in the ass.
I've been using a Scythe Grand Kama Cross 2 inside a NZXT Phantom 530 since 2013. The main reasons were the low rpm 14mm fan at the top, which is really silent and does the job just fine (i5-4670k) and the RAM clearance. Oh, and it looks cool!
Excellent Video. Interesting finding. I guess downdraft is dead for regular cases.
5:07 any cooler is compatible if you have a dremel tool.
11:01 that moment when you know there's others having in mind the same question and the prove :) thank you
So I will share idea for that housing - Cooler Master HAF XB EVO. Very old, but perfect for DARK ROCK TF 2 and easy to rebuild front panel. Cool case for DIY project with TF 2
Yes. I can't deny I like my fan being pointed towards the side of a case. That's why I kelp the stock and cooler for now.
Please redo this test with an sff case like the cooler master nr200p thank you.
Thanks Mike!
I use a Dark Rock TF updraft in a CM HAF XB box case. Dark Rock TF has 2 fans. Updraft setting is still push-pull. And lower fan is low enough to suck hot air from motherboard. HAF XB has a big mesh just above CPU and GPU. Both CPU and GPU heat will directly go up, out of case. I use DR TF on a 5900X, good for gaming and light video editing. IMO, downdraft coolers are for box or horizontal cases, sucking cool air from above or push hot air upward. They are not for tower cases.
I specifically chose a downdraft cooler for a retro build that had a fan on the side panel, which reduced clearance and complimented the cooler well, but for a modern full-sized system like my daily driver, I’ll choose a tower cooler. I'll get whatever is the right tool for the job.
Hell yeah, give me that CPU cooler test
Great video Mike! Although, what would the downdraft coolers look like if there was ventilation from the side panel? I would think that'd have a huge advantage to them over the tower coolers. Mind you, most cases have gone with solid side panels now, so might be hard to find newer ones that do hehe
I noticed you had images of an open test bench, does that mean you have another video coming out to see which ones just cools better overall?
Great question. It really depends on how this content does TBH. The main problem we have is a real lack of availability of cases that support higher tower coolers while also having side ventilation.
I would suggest to use NR200 with vent side panel. It has decent CPU cooler clearance.
I remember watching few other channels tested both and tower coolers still performed better at every test.
@@grintcoalways Tower coolers perform better until you add at least one fan to the side bracket. Ideally dual 140s.
I agree that the downdraft coolers were greatly handicapped by the solid side panel. I think it would be more relevant if tests were also conducted on an open bench to measure the real potential of each cooler and then tested in the specific case for comparison.
It is always nice to see P-ATX in the video ;)
i've noticed that the direction of the fin array when installed can have an effect on airflow. having them positioned horizontally instead of vertically seems to help exhaust air out of the side instead of down towards the hot gpu where it will clash with the air being pushed up from there and possibly create a lot of turbulence in your case.
The test setup is skewed against downdraft coolers, you need to move the case exhaust fan to the top (perpendicular to heatsink fins) for downdraft coolers to be a little more efficient. Some heat is being recirculated in the case. Best would be to have a side intake fan if possible.
Anyway, downdraft coolers have a good use for vented side panels like NR200, NCase M1, Cougar QBX, etc. They perform better with fresh intake directly going to the heatsink & components usually get 5-10°C lower temperatures. Just make sure you have an exhaust fan so heat doesn't get recirculated inside the case.
Not sure if your still looking at comments or not. Looking back at this video I see a flaw in the testing that you touched on, down draft coolers are not designed to work in modern cases due to the direction of air flow.
A more "accurate" test for direct cooler comparisons would be to use both on an open test bench and see how they directly effect temps. This would remove the "advantage" you gave the tower coolers with the use of a case optimized for their air flow pattern.
I have the Dark Rock TF with two fans exhausting air out the top of the case which is also the way the fins on the CPU cooler are aligned. I think this would make a noticeable difference in temperatures compared to the setup in this test. Having a fan at the back of the case favours the tower style cooler because that is the way the fins are aligned but the fan at the back of the case would do very little for a downdraft cooler if the fins are aligned to the top. The Dark Rock TF is complete overkill for my system which is Be Quiet themed and I mostly got if for the looks but I think having fans at the top of the case is the way to go with this kind off CPU cooler if the fins are aligned that way.
High-end motherboard isn't really the best choice to have conclusive results on VRM temperatures.
It would be much better to have such a test done on a mid-tier or garbage motherboard with a VRM that heats up to at least 80 degrees - that's where helping to cool the VRM starts to make a difference.
I can add my 2 cents into this conversation about both coolers, the dark rock 4 and dark rock tf2 as I had both installed (currently sticking to TF2). I have a Mini cube case with a single 140mm fan on the front, 120mm on the back and a chance to add a 120mm top fan as well. At the front I have Noctua Industrial 3000pwm set to 30% under 80 C and it goes up from there as required (I don't mind the fan noise, I actually like the swoosh/white noise at 30-45% to have in background when I'm working). On the back I have Redux 1700pwm set to silent. My motherboard lies flat in the case therefore with the TF2 I can access top airflow. My specs are Ryzen 7 5800X3D, B555M steel legend. Both coolers were configured manually to best performance. When I was running Dark Rock 4 my CPU temps at 105W was around 81-82 degrees under full load stress test after 30 minutes. VRM temp was around 60-61 C and ram 39-41 C. When I switched to TF2 CPU temps at 105W on my CPU are now topping a 78 (highest one I've noticed under the same stress test for the same amount of time), VRM temps dropped to 48(!!!) C and Ram temperature went down to 34C. I have actually noticed a drop in GPU (5700XT) temp by 3 degrees when using TF2. It really depends a lot on the case and how you can utilise airflow, and in case of TF2 when I added a extra top fan (noctua redux 1700) the temps dropped by about 2 degrees more but to keep it long term its a bit bad due to clearance issue, Soon I'm actually going to be upgrading the GPU to 7900XTX which I have tested before on both coolers and under heavy gaming in the same conditions I have also noticed nearly 5 degrees difference when using TF2 instead of Dark Rock 4. In my conclusion, if you do have a case that can utilise airflow from the top, it can really be beneficial as the air will actually cool your VRM and Ram and potentially GPU as well depending on your configuration. Overall, TF2 is an excellent cooler but I would say its more oriented for cases where the fans can get direct access to airflow.
Good video. I like a downdraft for looks as well. The AMD Wraith Prism is a great looking cooler even thought it's not that great of one.
Anyone remember those cases from years ago with side intakes that had actual plastic funnels leading to the CPU? Was probably back in the days when downdraft coolers (like the Zalman 'flower') were commonplace.
Feels like cases today could learn lessons from those days, even if we weren't running 190 Watt CPUs (much) back then :)
yeah. 80mm /120mm fan on the acrylic side. those can cool the Athlon XP's and P4's back then.
They don't look great but certainly helped a lot. Plus the PSU acts as vents for these coolers as it is located at the top.
I would like to see what an optimized case and case fan setup for the down draft or up draft cooler design would be.
With so few cases these days sporting mesh side panels, not sure what an "optimal" case would look like that would support all these coolers for normalized testing.
In addition to a classic mid tower air flow I have a big 500rpm 1700mm Noctua fan inside the tower door. So it sucks in additional cold air slowly on the whole board from basically up the mainboard, while there is a constant "classic" flow from the front to back. I could imagine in my case a down draft cooler might make a lot of sense and will be efficient.
What I dont like about dense fins is that it gets easier blocked by dust over time though. If u dont clean it, over the years there will be a layer of dust almost completely blocking all the air flow towards the CPU :)
Oh and I heard m.2 SSDs work better when under some heat..So that might be a + for downdraft coolers too..
The higher temperatures on the VRM and memory you've measured with downdraft coolers is probably due to 2 factors: hot air from the CPU being blasted over them (as you mentioned) but also the fact that downdraft coolers disrupt the "normal" airflow in a case. By the looks of it, you used 2 intake fans at the front and 1 exhaust on the back, pretty standard and it gives you a front to back airflow. The downdraft cooler gives a lot of turbulent air in that flow, not helping cooling.
I generally would only use downdraft coolers in 2 situations
1) Home theater pc's which you want to keep as quiet as possible, thus eliminating as many fans as can be. When using an APU, the CPU cooler alone should be enough for the entire system (the rest being passive exhaust) and in that case some extra airflow over the VRM can help, certainly on lower tier motherboards who sometimes lack VRM heatsinks
2) Office pc's where, just as with the home theater pc, you don't want extra fans in the case.
A third option would be if you want a specific downdraft cooler for looks, but then you have to except the drawbacks.
I constantly had to stare at that empty exhaust fan spot at the top 🤔
Backside GPU Temp differences, expectations RTX3090 rear mem temps...ah nope. Well done vid tho 👍
I had a Scythe tower cooler and I switched it to AMD prism cooler just to get VRM temps down. I have a Ryzen 5600 on a A320 motherboard so yeah. VRM temps measured from HWinfo dropped from 110 C to 75C! Test was Cinebench R23.
I have been using the Original Dark Rock TF since 2016 and I love it both in looks and performance and I've been waiting patiently to see a test like this. I only wish it had a Am4 bracket.
I feel like thisvideo made a good case for not using a downdraft cooler in an atx build. Maybe shoot something about them vs aio's for itx builds.
Absolutely. Like I said, this is setting a foundation to build more content on.
@@HardwareCanucks Looking forward to it
you talked so much about the importance of a case and then went with the one that greatly favours the tower
If only there existed some kind of visual tool for summarizing the numerical results in one image to allow for comparisons to be made more efficiently.