150mm maximum Tall Air Cooler recs?

Scythe Fuma? 120 mm dual tower. Shows it at 149 mm. I have one and I like it. Cheap too. Sub $50 ( at least it was when I bought it).
 
Scythe Fuma? 120 mm dual tower. Shows it at 149 mm. I have one and I like it. Cheap too. Sub $50 ( at least it was when I bought it).

I'm working on a ITX board. The width is not really realistic for dual towers. The case is only 15" long, 13" wide, 13 tall. and that's OD. For instance, my 240mm CLC can't be mounted int eh front unless I don't want to mount front bottom and front top fans. Just no room. If I mount it in the top, then I can mount front and bottom fans x4. The Noctua L12S covered all of the RAM, VRMs, and almost the entire board. And it's a 120mm top down riser cooler.
 
That's 63.00 USD. Fat chance. Cryorig can cry. Plus it's just your basic tower cooler that's been around for a couple decades. Noctua needs to step up its aesthetic game too. It's coolers are fugly as hell. The exceptions are both of these:

Noctua NH-L9a

View attachment 226725

Noctua NH-L9x65
This one doesn't come in black:
View attachment 226726

Those are nice, tight looking coolers. Here're the pros and cons with both of them.

Noctua NH-L9a
+Comes with longer fan screws to use their 92x25mm fan, which puts out 56 m³/h vs 79 m³/h. I don't know if that would increase cooling efficiency or not. I can't find any tests. If it could improve cooling by 10C at 100% CPU load, I'd be done with my search.
+Comes with screws, instead of clips, to mount either the slim or regular fan.
+Looks great, fits all systems as it confirms to AMD's stay out guidelines.
-Doesn't cool any better than the stock AMD stealth cooler using the slim fan. (Although, it is much quieter.)
-No tests with fat fan.
-No fat fan option, so you have to buy it, increasing cooler price into high end price level.

Noctua NH-L9x65
+Looks good
+Marginal cooling, but cools better than the AMD Spire by 4C, but worse than the AMD MAX/Prism by 2C. Test
+Much quieter than the AMD stock coolers (30dbs at 20CMs) vs 36 for the AMD MAX (every 3 decibels increases sound x2)
-No black version of the cooler. (Color of the regular Noc fans reminds me of my dentist office.)
-No fan screw mounting, relying on clips.
-Supposedly no alternate clips for the fatter fan, but you can get the 92mm black version of the fan in either 14 or 25mm version. I'm pretty sure teh fatter fan would increase cooling efficiency. But if they do not offer the clips to use the fatter fan, then it's zip tie time.
-No fat fan option, so you have to buy it, increasing cooler price into high end price level.

If I could find some fat fan specs on the Noctua NH-L9a, and it could do 5-10C better than the stock fan, I would go with that option. I'd probably go with teh L9x65 if I could find more cooling information on that, too. I'd just buy the black fan and be done with it.

--Another problem: Availability of the black 92mm fan in 92x25mm !!!
The fatter fan mounts fine on the L9x65. Just use the other mounting position for the clips (they are mounted in the lower position for the thin fan, in the picture). I have one, and I have used a 92x25mm fan with it.
 
I'm in the process of putting my new rig together. I need a Cooler for the AMD R5 3600 CPU - I will not use that shitty stealth cooler AMD ships with it. My case has a maximum limit of 150mm for the cooler.

I can't seem to find a decent looking cooler for a decent price that actully cools decently - lol.

Any ideas?

(I don't want to use the AMD Prism cooler either -too noisey)
If you've got some cash, the Dark Rock TF from Bequiet! is a solid cooler. And maybe the best for the height.
https://www.bequiet.com/en/cpucooler/572
 
I'm working on a ITX board. The width is not really realistic for dual towers. The case is only 15" long, 13" wide, 13 tall. and that's OD. For instance, my 240mm CLC can't be mounted int eh front unless I don't want to mount front bottom and front top fans. Just no room. If I mount it in the top, then I can mount front and bottom fans x4. The Noctua L12S covered all of the RAM, VRMs, and almost the entire board. And it's a 120mm top down riser cooler.
Based on this image, I think the Fuma would be just fine. Its similar in dimensions to the Mugen 5, which I have crammed into a Raijintek Metis. it doesn't hang over the top of the board. You gotta be careful, though. The Fuma 2 is out and its 154mm tall.
1586754385771.png
 
Based on this image, I think the Fuma would be just fine. Its similar in dimensions to the Mugen 5, which I have crammed into a Raijintek Metis. it doesn't hang over the top of the board. You gotta be careful, though. The Fuma 2 is out and its 154mm tall.
View attachment 237579
Hey we have the same case. If I may ask, how do you keep your card from drooping?
 
If you've got some cash, the Dark Rock TF from Bequiet! is a solid cooler. And maybe the best for the height.
https://www.bequiet.com/en/cpucooler/572
Well, the Noctua L12S would have been fine too, but it was huge vs my board, and covered the RAM too. I'd have to do the measurements on that to see if it would even fit. However, I don't need that much cooler. ITX board clearance is tight. Some DLED plugs even have to be trimmed to fit on the DLED riser due to space constraints on those boards. I wish I had just tried the Noctua L9A first. I just had a feeling it was going to throttle on the 3600. Still should have installed it and just tested it first.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
I hope those will work on my case. On the Corsair Crystal 280x, the PCI card lock down is on the outside of the case. There are four possible expansion slots, but they lock down on the outside with a clamp like design and two case screws. There is no way to screw down the support itself :( There aren't any threads in the case. If the brace adds any extra width to the PCI cards attachment, it may not clamp correctly, but I'm gonna try it.

LOL, yeah, need the RGB. Ironically, this build was going to be static in black, white, and grey only, and only a single color RGB light to illuminate it from teh top down. And then. . . .
 
Last edited:
As an Amazon Associate, HardForum may earn from qualifying purchases.
Little late, but oh well.
What motherboard do you have?
While Scythe Fuma is short enough (137x131x149mm (WxDxH) front of cooler is 50mm center CPU to front of fins & 51.6mm to back of fins and performs quite well with ony middle fan (only 1-3c warmer than 2x fans).
 
Little late, but oh well.
What motherboard do you have?
While Scythe Fuma is short enough (137x131x149mm (WxDxH) front of cooler is 50mm center CPU to front of fins & 51.6mm to back of fins and performs quite well with ony middle fan (only 1-3c warmer than 2x fans).

MB: Gigabyte Aorus Pro Wifi ITX.
I'm running the Enermax Aquafusion 240 in white. I still like air better because it is more simple than liquid cooling, which jsut adds more moving parts, and liquid, which is never good with electronics. You're not too late. I won;t be the only person with this problem, obvisouly, as you can see maniufactureres like Noctua trying to market the small form factor cooling problem.
 
Which model Aorus? Should be something like "B450 AORUS Pro Wifi (rev. 1.0)"
 
Okay, if you are in front of Asus B450 Aorus Pro Wifi (rev. 1.0) ITX with motherboard I/O farthest from you, to your left you have almost 90mm center CPU to near side of PCIe socket but only about 58.9mm to the right of CPU to side of motherboard. You have about 51.8mm center CPU toward you to near side of RAM socket and about 83.5mm to back of motherboard over I/O panel. Cooler width will likely be dependent on how far fins reach from center CPU toward case on right side and back of case and of course it will depends on how tall RAM is. The only direction with no issues is toward PCIe socket.
 
Okay, if you are in front of Asus B450 Aorus Pro Wifi (rev. 1.0) ITX with motherboard I/O farthest from you, to your left you have almost 90mm center CPU to near side of PCIe socket but only about 58.9mm to the right of CPU to side of motherboard. You have about 51.8mm center CPU toward you to near side of RAM socket and about 83.5mm to back of motherboard over I/O panel. Cooler width will likely be dependent on how far fins reach from center CPU toward case on right side and back of case and of course it will depends on how tall RAM is. The only direction with no issues is toward PCIe socket.
Aorus is Gigabyte. Yep, exactly what I had to do when looking. However, I found that even when a cooler "clears" it can be problematic. The Noctua L12S, for instance, covered the RAM, even though it "cleared" it. You can't take out the inner RAM bank unless you take off the cooler. What I should have done, since it is easy to bend, made out of soft copper, is just bend the radiator up and left it at a 30-45 degree angle. That would have given access to the RAM bank, and may have looked pretty cool too. I don't think it would have affected cooling performance at all. However, it's not a very nice looking cooler. It's nice from a technical POV and it's crafted with precision in all ways. In that respect, the Noc L12S is VERY nice looking.

I still wish I had tried the L9A. It probably would have been fine, and it's clean as hell looking installed in a board like an ITX (i.e., even though it is 92mm, ti looks fat in the ITX board. I could have added a 92 to 120 adapter to it and mounted a 120mm fan on it too. It could have been customized to look really nice. I contemplated having a custom shroud made from a 3D printer to go from the cooler, up over the videocard (going from round to long and skinnier, something like 92mm pipe then flattening out to 6" x 1/2", giving the same airflow dimensions, but clearing teh video card, then curving around the top of the video card and back down to the bottom case vents, where the shroud would open up again. A custom decal on the intake shroud would round it out.)
 
I've tested and used several downflow and many more tower coolers over the years. Downflow are problematic for keeping their heated exhaust air from mixing with their cool intake air, especially when pushing air toward motherboaard. Their airflows through cooler hitting motherboard and turning out where it hits RAM, GPU, etc turning up along side of cooler and fan where it is drawn right back in. End result is 8c higher temps because intake airflow can easily be pre-heated 7-9cc when CPU is working hard. I found this happened on open bench test station with 140mm fan flowing air over motherboard. The good news is simply turning fan over so it pulls air away from motherboard can solve the problem. But whenever possible I prefer using tower coolers. There are some very good ones out there, like Cryorig M9, Noctua NH-D9L, NH-U9S, NH-U9L, Thermalright TRUE Spirit 90M, and others. All it takes is getting a case with enough clearance for a small tower cooler instead of a downflow cooler.
 
I've tested and used several downflow and many more tower coolers over the years. Downflow are problematic for keeping their heated exhaust air from mixing with their cool intake air, especially when pushing air toward motherboaard. Their airflows through cooler hitting motherboard and turning out where it hits RAM, GPU, etc turning up along side of cooler and fan where it is drawn right back in. End result is 8c higher temps because intake airflow can easily be pre-heated 7-9cc when CPU is working hard. I found this happened on open bench test station with 140mm fan flowing air over motherboard. The good news is simply turning fan over so it pulls air away from motherboard can solve the problem. But whenever possible I prefer using tower coolers. There are some very good ones out there, like Cryorig M9, Noctua NH-D9L, NH-U9S, NH-U9L, Thermalright TRUE Spirit 90M, and others. All it takes is getting a case with enough clearance for a small tower cooler instead of a downflow cooler.

They have the benefit of cooling the VRMs, too, though, whereas the towers do not. But I agree that the tower is more efficient for that reason. To get around recirculating air, at least somewhat, you can use a fan shroud, or riser, that bolts right onto the fan or radiator, raising it up an inch or more. Noctua makes a monster top down cooler that avoids that problem becasue the fan is so far above the MB.
Fan Shroud
Noctua huge top down:
81SPOa8QnuL._AC_SL1500_.jpg
This thing is like 150mm tall with the fan on top.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
They have the benefit of cooling the VRMs, too, though, whereas the towers do not. But I agree that the tower is more efficient for that reason. To get around recirculating air, at least somewhat, you can use a fan shroud, or riser, that bolts right onto the fan or radiator, raising it up an inch or more. Noctua makes a monster top down cooler that avoids that problem becasue the fan is so far above the MB.
Fan Shroud
Noctua huge top down:
View attachment 238273
This thing is like 150mm tall with the fan on top.
That cooler is still dumping heated air in all directions once it hits motherboard.

If case has decent airflow and RAM isn't too tall VRM cooling is not an issue. Even my old i7 920 @ 4.3 GHz which made VRM very warm cooled just fine with tower coolers. Fan/s on tower cooler push air under cooler anyway .. assuming RAM doesn't have super tall so-called "heatsinks" that don't lower RAM temps anyway, but force cooler fans to be mounted higher thus blocking airflow over motherboard .. also a problem with even the cooler you posted. A tower cooler draws air in it's front and pushes heated air out it's back directly into rear vent where it leaves the case .. thus it doesn't mix with and heat up air entering cooler. Even your posted cooler pushes air down onto motherboard where the spreads in all directions hitting GPU and even short RAM causing it to turn up the sides of cooler and mix with cool air going toward cooler pre-heating it before it goes into cooler.

Key to cool quiet system is keeping heated cooler exhaust separate from intake air so it does mix with it and make it warmer going into coolers. Every degree warmer air enter cooler is means same degrees hotter that component will be at same fan speed.
 
That cooler is still dumping heated air in all directions once it hits motherboard.

If case has decent airflow and RAM isn't too tall VRM cooling is not an issue. Even my old i7 920 @ 4.3 GHz which made VRM very warm cooled just fine with tower coolers. Fan/s on tower cooler push air under cooler anyway .. assuming RAM doesn't have super tall so-called "heatsinks" that don't lower RAM temps anyway, but force cooler fans to be mounted higher thus blocking airflow over motherboard .. also a problem with even the cooler you posted. A tower cooler draws air in it's front and pushes heated air out it's back directly into rear vent where it leaves the case .. thus it doesn't mix with and heat up air entering cooler. Even your posted cooler pushes air down onto motherboard where the spreads in all directions hitting GPU and even short RAM causing it to turn up the sides of cooler and mix with cool air going toward cooler pre-heating it before it goes into cooler.

Key to cool quiet system is keeping heated cooler exhaust separate from intake air so it does mix with it and make it warmer going into coolers. Every degree warmer air enter cooler is means same degrees hotter that component will be at same fan speed.
I'm not disagreeing with the theory, but every review I could find on that cooler reported excellent cooling. If you have good case airflow, recycling air isn't goign to be a probem. "The comparatively huge mass of the cooler and the excellent NF-A14 fan offer the NH-C14S a tremendous performance advantage. As a matter of fact, the NH-C14S is a threat to many large tower coolers, outperforming the Cooler Master EVO 212 by a significant margin and touching the performance of the Grandis and the Dark Rock Pro 3, all while maintaining reasonably low noise levels." ANADTECH TWEAKTOWN
 
I'm not disagreeing with the theory, but every review I could find on that cooler reported excellent cooling. If you have good case airflow, recycling air isn't goign to be a probem. "The comparatively huge mass of the cooler and the excellent NF-A14 fan offer the NH-C14S a tremendous performance advantage. As a matter of fact, the NH-C14S is a threat to many large tower coolers, outperforming the Cooler Master EVO 212 by a significant margin and touching the performance of the Grandis and the Dark Rock Pro 3, all while maintaining reasonably low noise levels." ANADTECH TWEAKTOWN
I'm not saying flat coolers can't cool well. I'm only saying they make it harder to get good temps than tower coolers. Every degree warmer the air is entering cooler translates into same degrees hotter componet (CPU will be at same fan speed. If air entering cooler is 10 c warmer than room the CPU will also be 10c hotter at same fan speed. I've seem custom ordered and assembled systems that had air 20c warmer than room entering CPU and GPU coolers when gaming. These were built by companies who supposedly knew hwat they were doing but obviously didn't know anything about setting up case airflow properly. Fact is most builders don't understand and implement good case airflow.

Recycled heated exhaust is the biggest cause of high temps and high noise systems, which is why those of us who understand airflow do as much as we can to stop it. The best way to use flat coolers like NH-C14S is with fan moving air away from motherboard. As for them being "a threat to many large tower coolers" that is pure fiction .. and EVO 212 is definitley not one of them. Any quality120mm fan cooler is easily as good and usually better than EVO 212, and all decent140mm fan towers stop it into the mud. EVO 212 was/is a mid-size budget cooler that used to be one of only a few good value coolers, but for serveral years now we have had many others that perform same or better costing simliar or less.

Anadtech testing do a decent job of testing coolers using a fixed heat source and monitoring/recording air temp into cooler at same time CPU temp is taken. But we have to understand air temp into cooler is as critial to cooling as having a good cooler and setup our case airflow accordingly.
Tweaktown's test procedure doesn't say anything at all about air temp, ambient temp. My guess is they use room ambient which can easily be +/- 3-5c of actual air temp entering cooler, usually higher becuase of other heat sources around test station .. like PSU, lighting, even heat coming off of people doing testing.

If you are interested in finding out how hard it is to test coolers, take a look at "Why Most Cooler Tests are Flawed: CPU Cooler Testing Methodology".
https://www.gamersnexus.net/guides/3561-cpu-cooler-testing-methodology-most-tests-are-flawed
 
I'm not saying flat coolers can't cool well. I'm only saying they make it harder to get good temps than tower coolers. Every degree warmer the air is entering cooler translates into same degrees hotter componet (CPU will be at same fan speed. If air entering cooler is 10 c warmer than room the CPU will also be 10c hotter at same fan speed. I've seem custom ordered and assembled systems that had air 20c warmer than room entering CPU and GPU coolers when gaming. These were built by companies who supposedly knew hwat they were doing but obviously didn't know anything about setting up case airflow properly. Fact is most builders don't understand and implement good case airflow.

Recycled heated exhaust is the biggest cause of high temps and high noise systems, which is why those of us who understand airflow do as much as we can to stop it. The best way to use flat coolers like NH-C14S is with fan moving air away from motherboard. As for them being "a threat to many large tower coolers" that is pure fiction .. and EVO 212 is definitley not one of them. Any quality120mm fan cooler is easily as good and usually better than EVO 212, and all decent140mm fan towers stop it into the mud. EVO 212 was/is a mid-size budget cooler that used to be one of only a few good value coolers, but for serveral years now we have had many others that perform same or better costing simliar or less.

Anadtech testing do a decent job of testing coolers using a fixed heat source and monitoring/recording air temp into cooler at same time CPU temp is taken. But we have to understand air temp into cooler is as critial to cooling as having a good cooler and setup our case airflow accordingly.
Tweaktown's test procedure doesn't say anything at all about air temp, ambient temp. My guess is they use room ambient which can easily be +/- 3-5c of actual air temp entering cooler, usually higher becuase of other heat sources around test station .. like PSU, lighting, even heat coming off of people doing testing.

If you are interested in finding out how hard it is to test coolers, take a look at "Why Most Cooler Tests are Flawed: CPU Cooler Testing Methodology".
https://www.gamersnexus.net/guides/3561-cpu-cooler-testing-methodology-most-tests-are-flawed

I'm not disputing anything you said.

All tests are flawed unless you have an agreed upon method of testing with clear controls. That's why we have to test drugs so rigorously, which have generally preclinical (animal) tests, then phase 1-3 clinical trials (humans) that take of years testing for safety Phase1, Safety and efficacy Phase 2, then safety and efficacy again during a large Phase 3 trial. It's all about controls.

The reason most cooling towers, industrial and computer, are vertical is because that is a more efficient design. Because of physics, vertical coolers will be more efficient. Recycled air can be dealt with.

My point is that all of the Noc top down coolers from the L9a to the large 140mm job do a really good job of cooling - given their intended purpose.
 
I'm not disputing anything you said.

All tests are flawed unless you have an agreed upon method of testing with clear controls. That's why we have to test drugs so rigorously, which have generally preclinical (animal) tests, then phase 1-3 clinical trials (humans) that take of years testing for safety Phase1, Safety and efficacy Phase 2, then safety and efficacy again during a large Phase 3 trial. It's all about controls.

The reason most cooling towers, industrial and computer, are vertical is because that is a more efficient design. Because of physics, vertical coolers will be more efficient. Recycled air can be dealt with.

My point is that all of the Noc top down coolers from the L9a to the large 140mm job do a really good job of cooling - given their intended purpose.
Indeed, we both understand the problem.

I'm not sure what you mean by "The reason most cooling towers, industrial and computer, are vertical is because that is a more efficient design ". Assuming we are only talking about heatpipe coolers which are the vast majority of computer coolers, heatpipes work so close to the same horizontal and vertical it's hard to measure a difference, but tower coolers definitelly do make it much easier to deal with heated exhaust.

Indeed, there are many very good top down (or bottom up) coolers, but I'm still going to pick cases that have enough room for tower coolers instead of fighting heated exhaust. That said, my SST-RVZ01 HTPC has Thermalright AXP-100 in it with vent directly over it. :facepalm:
 
Last edited:
They have the benefit of cooling the VRMs, too, though, whereas the towers do not. But I agree that the tower is more efficient for that reason. To get around recirculating air, at least somewhat, you can use a fan shroud, or riser, that bolts right onto the fan or radiator, raising it up an inch or more. Noctua makes a monster top down cooler that avoids that problem becasue the fan is so far above the MB.
Fan Shroud
Noctua huge top down:
View attachment 238273
This thing is like 150mm tall with the fan on top.

Thats a 140mm monster, not sure why you would use it , when the 92mm towers out-perform it for less mass/money. It seems to me the only time to not use a tower, is when you absolutely can't even fit a 92mm tower based cooler.

13084356127l.jpg

13084030641l.jpg
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
Indeed, we both understand the problem.

I'm not sure what you mean by "The reason most cooling towers, industrial and computer, are vertical is because that is a more efficient design ". Assuming we are only talking about heatpipe coolers which are the vast majority of computer coolers, heatpipes work so close to the same horizontal and vertical it's hard to measure a difference, but tower coolers definitelly do make it much easier to deal with heated exhaust.

Indeed, there are many very good top down (or bottom up) coolers, but I'm still going to pick cases that have enough room for tower coolers instead of fighting heated exhaust. That said, my SST-RVZ01 HTPC has Thermalright AXP-100 in it with vent directly over it. :facepalm:
It really doesn't matter what type of cooler you're talking about, passive, heat pipe,water cooled gas, etc., it's just the physics of heat that dictate the vertical configuration.
Good place to start I guess: https://en.wikipedia.org/wiki/Cooling_tower
 
Thats a 140mm monster, not sure why you would use it , when the 92mm towers out-perform it for less mass/money. It seems to me the only time to not use a tower, is when you absolutely can't even fit a 92mm tower based cooler.

View attachment 239347

View attachment 239348
Link? The links I provided show a different story. This is really just physics. The more cooling area you have passing xx% cooler matter through it, be that gas or liquid, the better cooling you'll get--unless there is a mechanical flaw in the cooler itself. One reason to go with a 140mm cooler is that you can pass the same amount of air through it at a much slower fan speed compared to a 120mm or smaller fans. That's why Corsair PSUs went from 120 to 135mm fans.
 
It really doesn't matter what type of cooler you're talking about, passive, heat pipe,water cooled gas, etc., it's just the physics of heat that dictate the vertical configuration.
Good place to start I guess: https://en.wikipedia.org/wiki/Cooling_tower
You need to study up of heat transfer in heatpipes, solid mass, etc. Heat does not rise all by itself. Heatpipes move heat using phase change. The boiling of water at heat source expands and pushes previou phase change toward ends of heatpipes. Modern heatpipes move heat towar ends at almost exactl the same rate regardless of orientation. Read the link you posted. Maybe then you will understand what cooling towers do and why it's not the same as our computer cooling.
 
Link? The links I provided show a different story. This is really just physics. The more cooling area you have passing xx% cooler matter through it, be that gas or liquid, the better cooling you'll get--unless there is a mechanical flaw in the cooler itself. One reason to go with a 140mm cooler is that you can pass the same amount of air through it at a much slower fan speed compared to a 120mm or smaller fans. That's why Corsair PSUs went from 120 to 135mm fans.

The links you show don't appear to be in a conventional computer case.

That paints an erroneous rosy picture of top down coolers, better than the typical reality of putting them in a conventional case, where a tower can direct the majority of the heat out of the case, while a top down cooler recirculates the heat.

That explains the difference.

The test I linked from used a conventional case:
https://www.overclock3d.net/reviews/cases_cooling/cpu_air_cooler_mega_test/1
 
Last edited:
Link? The links I provided show a different story. This is really just physics. The more cooling area you have passing xx% cooler matter through it, be that gas or liquid, the better cooling you'll get--unless there is a mechanical flaw in the cooler itself. One reason to go with a 140mm cooler is that you can pass the same amount of air through it at a much slower fan speed compared to a 120mm or smaller fans. That's why Corsair PSUs went from 120 to 135mm fans.
You make it sound like all you need to cool is lots of airflow with no regard to how warm or cool that air is. Reality is our coolers respond at almost a 1:1 ratio of air temp into cool to CPU temp at given load and airflow rate (fan rpm). Increase air temp into cool 5c and CPU temp goes up similar 5c at same load and fan speed. Put a high powered GPU into the mix in a case with poor airflow and a downflow cooler and the air inside of case and air entering cooler after gaming for 10-20 minutes can be 15-20c higher than in room.

As others and myself have repeately said, it's much easier to move heated exhaust air out of a tower CPU cooler moving on back out out of case than with a downflow cooler. Now if we could only get industry to market something similar for GPUs and things would be golden.

AS for PSU usign 120mm, 135mm or 140mm fans it doesn't make much difference. The use of more efficient PSU components make much cooler running PSUs than bigger fans do. ;)
 
Back
Top