The unsatisfying state of air cooling solutions on the market (Case/GPU/NT) - idea as pic

[firejohn]

The main problem is with the GPU spreading all the heat around the system components and also the CPU.
Blowers are loud, while Top-Blowers spread the heat into system.

So short of thoughts i painted this pic:

Also on GPU-top-blowers on the right side it blows the heat upwards the fan of the CPU (its fan sucks it to itself) so warm air goes through the heatsink of the CPU.
Also those GPU air solutions blows directly to the mainboard/PCIe, depends on the direction of the fins. And mostly those cards have it open in direction to the mainboard.

So in the years they started to put the NT on the bottom with its own airchannel in case manufacturing, they could do this too with CPU and GPU to seperate them and
create their own airchannels.

The heatsink of the GPU should be redesigned, fins into length, in direction from front to rear case and a higher profile. So on some GPU cards it would be possible to use them without
any fans attached to it and let the casefans do the work. Generally casefans are more quiet and more reliable.

Also the rear of the case should have a place under the GPU for a casefan to suck the warm air out.

The alternative NT design, just a small additional idea but not so important. Atm in my case for example theres a small space between fan and bottom (without holes), that is also not
optimal.

I hope some manufactures sees this or have their own ideas.

 

[rtwjunkie]

I’m pretty satisfied with my temps on my setup. Very satisfied in fact with my air cooling. None of my temps ever get anywhere close to worrisome.

 

[FordGT90Concept]

That's pretty much BTX:

BTX (form factor) - Wikipedia

en.wikipedia.org

 

[8BitZ80]

This is getting into the age-old "most efficient cooling solution" discussion, but I think you'll find that the most efficient cooling setup will exhaust out the top. Hot air rises and so an exhaust fan on top will draw all the hot air straight out.
Think about the garbage can 2013 Mac Pro: Terrible machine, but one thing it did brilliantly was it's almost-ideal cooling configuration. Intake on the bottom, exhaust out the top, that means hot air can be pulled straight out as efficiently as possible. That's what you want.

 

[mtcn77]

If you are into clockwise turned motherboards, silverstone raven 2 cases make that vertical alignment.
You will have to make due with blower type gpus since heatpipes don't correctly orient in the vertical axis. However for best performance we would need cooling top arrangement in both the gpu and the cpu anyway. So a custom benchtop with a riser will beat all else.

 

[TheoneandonlyMrK]

If you are into clockwise turned motherboards, silverstone raven 2 cases make that vertical alignment.
You will have to make due with blower type gpus since heatpipes don't correctly orient in the vertical axis. However for best performance we would need cooling top arrangement in both the gpu and the cpu anyway. So a custom benchtop with a riser will beat all else.

I agree with the Op , It would be nice to see some more outlandish cooling setups, but more extreme then him for me, I want triple phase chambers with a Tec on top then what most would consider a massive cooler on top of that, and that's just the CPU.
Wtaf not have a Four nah six card slot GPU cooler, who uses more than one slot , like 5% of us tops, put three vapour chambers in that with sixteen heeatpipes on top both cooled by some kind of hybrid jet effect fan ,ultra high pressure low rpm fans.
That's quite too

 

[mtcn77]

I think vertical is still the best for the case of laminar flow air impedance reduction. If we had back face sweeping large blowers I think we could reduce cooler noise a bit, however they are independable as load increases as flow decreases. That is the opposite with a forward blade blower fan. So we cannot have those small induction motors in the same unit hub.

 

[TheoneandonlyMrK]

I think vertical is still the best for the case of laminar flow air impedance reduction. If we had back face sweeping large blowers I think we could reduce cooler noise a bit, however they are independable as load increases as flow decreases. That is the opposite with a forward blade blower fan. So we cannot have those small induction motors in the same unit hub.

Hey I'm just pulling ideas out my butt, I'm not scienceing the crap out of this rig, but I have been getting fresh ideas on an external to my pc removable 360mm stood rad just cooling four recs, meant TECS that then drop my water temps to ambient fully, with the potential for lower.

 

[mtcn77]

Hey I'm just pulling ideas out my butt, I'm not scienceing the crap out of this rig, but I have been getting fresh ideas on an external to my pc removable 360mm stood rad just cooling four recs, meant TECS that then drop my water temps to ambient fully, with the potential for lower.

I heard you get better convection with some spacing in the middle. I'm most positive no one makes hub shrouds to the radiator dead zone where the motor sits.

 

[TheoneandonlyMrK]

I heard you get better convection with some spacing in the middle. I'm most positive no one makes hub shrouds to the radiator dead zone where the motor sits.

I plan to use an unconventional design, my old cases whole front door, Thermaltake Kandalf Lcs had ,well still atm has a swinging fully aluminuim front door, always liked it, with a weird 360 threaded pipe radiator, like a heatsink ,one pipe 10 passes, it worked well with a lot of air flow across it from push pull, i have phantec standoffs set asside for it and enough fans ,waterblocks(ex gpu)and the controller , i just need a pump and some Tecs, and quick release fittings err and pipe but thats it and i can start building.

 

[INSTG8R]

I plan to use an unconventional design, my old cases whole front door, Thermaltake Kandalf Lcs had ,well still atm has a swinging fully aluminuim front door, always liked it, with a weird 360 threaded pipe radiator, like a heatsink ,one pipe 10 passes, it worked well with a lot of air flow across it from push pull, i have phantec standoffs set asside for it and enough fans ,waterblocks(ex gpu)and the controller , i just need a pump and some Tecs, and quick release fittings err and pipe but thats it and i can start building.

I’m still using that case today. The pump won’t die and I just replaced the CPU block with am AC Cuplex Kyros that keeps my 4790K surppriimgly cool for something so restrictive and designed for Pentium 4s

 

[TheoneandonlyMrK]

I’m still using that case today. The pump won’t die and I just replaced the CPU block with am AC Cuplex Kyros that keeps my 4790K surppriimgly cool for something so restrictive and designed for Pentium 4s

I couldn't throw the case on the tip, I'm even removing the door sympathetically, could all be reversed, it's a good job I'm skint or I'd be surrounded by pcs crunching.

 

[INSTG8R]

I couldn't throw the case on the tip, I'm even removing the door sympathetically, could all be reversed, it's a good job I'm skint or I'd be surrounded by pcs crunching.

You couldn't throw it in the tip because it weighs enough to throw out your back...I actually reversed my door years ago it was a fun hassle.

 

[Flanker]

I would like to have something like a 2U motherboard, except one of the socket is for GPU, and they can both use big tower air coolers. But then, I have no clue how to integrate the GPU circuitry and GDDR.

 

[Fouquin]

I would like to have something like a 2U motherboard, except one of the socket is for GPU, and they can both use big tower air coolers. But then, I have no clue how to integrate the GPU circuitry and GDDR.

HBM and SXM3 is the answer to that.

 

[Bill_Bright]

The unsatisfying state of air cooling solutions on the market

I reject this statement completely. The entire premise is simply false!

With the exception of extreme overclocking (and exceptions do NOT make the rule), if you cannot achieve proper cooling with fans (and without excessive noise too) then it is YOU who are doing an unsatisfactory job at selecting components (case and fans) and setting up your case cooling!

Thanks to the ATX Form Factor standard, there are lots and lots of air cooling solutions on the market giving us lots and lots of very effective options to choose from. And that includes, BTW, OEM CPU cooling solutions for those CPUs that come with OEM supplied coolers.

 

[sneekypeet]

giving us lots and lots of very effective options to choose from.

In the mind of a cooling reviewer, I reject this statement.

CPU cooling has taken a nose dive in the last couple of years in the pursuit of silence and RGB. There hasn't been a cooler that can compete with the top of my CPU cooler charts in two to three years. CPU cooler manufacturers are now happy with mediocre results, rather than anyone trying to improve the performance. I too am unsatisfied with the state of current options!

Just because a cooler "works" does not mean satisfaction is part of it.

 

[dirtyferret]

I’m pretty satisfied with my temps on my setup. Very satisfied in fact with my air cooling. None of my temps ever get anywhere close to worrisome.

+1
same, my temps are cool (and my CPU is OC) and my fans stay silent.

 

[mtcn77]

Latest generation coolers have two distinct modes of operation, imo. One silent with standard fans and another high power mode with reference fans are installed.
Scythe Fuma & Ninja target specific noise levels and cannot overcome some threshold without running into elevated air pressure. They can offer great improvements, albeit not overcome its associated noise. Such conditionalities force newer generations to slim down to lower air impedance, imo.

 

[Bill_Bright]

In the mind of a cooling reviewer...

Just because a cooler "works" does not mean satisfaction is part of it.

"In the mind of a" certified master electronics technician, yes it does!

There seems to be this unrealistic belief these days that something is good or satisfactory only if it is the best, and anything less is unsatisfactory. That's simple hogwash! It may seem to be true when it comes to personal opinions. But that is NOT true, or realistic in the unbiased and technical world of electronics cooling. The Laws of Physics and thermodynamics are not based on personal opinion.

If the cooling solution keeps the CPU comfortably within its specified normal operating temperature range, then it "works" and that is "satisfactory"!

It is wrong - from a purely technical aspect - to assume cooler is automatically better. That is just more hogwash! Yes, proper cooling is absolutely essential, but it is NOT necessary to get the coolest temps possible. There is absolutely nothing to suggest a CPU running at 30°C will be more stable, perform better, or have a longer life expectancy than a cooler running at 50° or even warmer.

PLUS - and this seems to be something you chose to ignore in your reply to me - proper cooling is also a function of good case cooling too. If the case does not adequately extract the heat the CPU cooler tosses up, then of course, CPU cooling will be affected. But that takes me back to my original comment where I expressly stated that it is the user's responsibility to select the proper case and set up proper case cooling too.

Do note, SP, the OP was NOT just addressing CPU coolers but cases too.

 

[GlacierNine]

This is getting into the age-old "most efficient cooling solution" discussion, but I think you'll find that the most efficient cooling setup will exhaust out the top. Hot air rises and so an exhaust fan on top will draw all the hot air straight out.
Think about the garbage can 2013 Mac Pro: Terrible machine, but one thing it did brilliantly was it's almost-ideal cooling configuration. Intake on the bottom, exhaust out the top, that means hot air can be pulled straight out as efficiently as possible. That's what you want.

The effect of hot air rising is completely irrelevant as soon as a fan is involved. I really wish people would stop regurgitating this little bit of common nonsense.

 

[dirtyferret]

The effect of hot air rising is completely irrelevant as soon as a fan is involved. I really wish people would stop regurgitating this little bit of common nonsense.

100%, while it does rise it does so slowly and I can stick intake fans in a PC case roof and shove the hot air right out the vented bottom of a case if I want.

CPU cooling has taken a nose dive in the last couple of years in the pursuit of silence and RGB.

That may have more to do with more cache and cores under load and worse thermal transfer from the CPU to the cooler.

 

[sneekypeet]

It may seem to be true when it comes to personal opinions. But that is NOT true, or realistic in the unbiased and technical world of electronics cooling.

Oh, I see, only your opinion and perspective matter. Cases have fallen into the same trap too, but I guess that is from someone who has seen damn near all of them, and not just a select segment of the market.

But I digress, as you are always the one who is right on all topics.

 

[Vario]

The effect of hot air rising is completely irrelevant as soon as a fan is involved. I really wish people would stop regurgitating this little bit of common nonsense.

I have top intake bottom exhaust, necessity on this case due to front fan (in the 3x3.5 cage) having to be intake to remove power supply exhausted hot air from the poorly ventilated front bezel, that makes the rear fan exhaust to avoid pressure issues, and therefore the top must be intake. No issues. Just have to have a reliable dual ball bearing design if running impeller up.

 

[freeagent]

Brute force with big fans. Quiet fans let you overheat. Small coolers will too.