If bottom is used as intake you would just dump more heat in the intake air when you use it between components or straight after them which would then in turn impact the performance of the top radiator which is exhausting.
Why not directly exhaust most of the heat from the top before going to the bottom intake rad to cool even further.
And thanks for the nickname.
Yes, that is how it "feels" isn't it? It seems logic.
However, that is not how it works. Unless the loop temperature is waaay hot, the difference between bottom & top rad will be a couple degrees at best. As a whole, the loop will try to get as close to ambient temp as physically possible. And while differences before/after block & rad exists, they are/should be minimal. And should not matter.
IF the water in your loop reaches the temp of the GPU (80°C) when it leaves the GPU block, then there IS a problem for the CPU, as the CPU will not like 80°C very much. However, the problem can not be resolves by routing the water through a radiator before it reaches the CPU. Unless it is a REALLY BIG radiator, it wont be able to drop the water temp enough to be "safe" for the CPU.... because AIR is limited in what it can "pick up".
However, the point is moot, because if you have a situation where the water can reach 80°C... you have other, more pressing problems as it just means that you dont have enough radiator/fannage. The loop is dimensioned wrong for the heatload.
Say that you have 300 Watts load (just add max tdp of CPU & GPU), then you should be looking at "enough radiator" that can dissipate at least 300 Watts at a given fanspeed WHILE the temperature does not rise more as 10°C above ambient. These are not easy calculations.. its easier to just look at all the charts Martin has created
If you have enough raddage/fannage, then the loop temp as a whole should not exceed ambient +10°C (+15°C) And certainly not the max temp of the most "fragile" element you are cooling (this might be the CPU , but it might just as well be the pump or the rubber gasket).
In other words, your loop should not exceed 50°C... ever.
The "bigger" the loop (read: the more/larger/better radiator(s)surface), the closer to ambient it will run.
In normal circumstances, like ambient 25°C and loop 35°C to 40°C even fluctuations of several degrees (iow the "hotter" water coming from the GPU(s)) will still be within the "safe" zone of the components.
Now, if you were living in some desert where the ambient temp is +50°C , then you would need A LOT, really a LOT of radiator surface (in the shade) to keep everything in the safe zone. But frankly, for cooling a PC in a +50°C desert there are better solutions.. evaporative cooling for example.
Anyways. Yes, the water coming from the GPU's hitting the CPU will be warmer than it would be if it went through a radiator first. But it should still be in the CPU "safe" zone. Actually, it WILL be in the safe zone, as it will only be a couple degrees warmer as the loop as a whole
If the loop is at ambient +10°C, then the water coming from the GPU might be ambient +15°C. But not much more, unless you've got waterblocks in unobtainium. If the water coming from the GPU's it "too hot" for the CPU, then it measn that the water was already too hot before in went into the GPU block - meaning not enough radiator.
Thus the loop order doesn't matter that much as long as reservoir is before and higher as the pump.
As for the radiator(s), it doesn't matter that much if they are before or after GPU or CPU or between GPU & CPU; as long as there is "enough".
Although i would avoid putting the radiator between pump & GPU/CPU.. but for different reasons.
As for "hot" air coming from one radiator being sucked through another radiator.. yes, it will have an impact on loop segment temps... but not as much as you would think.
It only means that the 2nd radiator will be less efficient because of the smaller air/water delta.
The loop will not run hotter than a loop which has only one radiator (which has the size/surface of the two seperate rads combined).
If it does, it means that there wasn't enough air "sucked in" to begin with.
And it even doesn't matter if the first rad is twice as "big" with twice as many fans as the second or vice versa.
It's counterintuitive , yes, but there it is.
On a positive note however, you can run the loop cooler by having both radiators taking in fresh/cool air from outside the case.
What? But my vrms & ram will melt! the air inside the case will run very hot!
No, they wont. Yes the air in the case will be a bit warmer, but nowhere near danger levels.. Unless your radiators are unobtainium and your loop "too small" to begin with.
Conclusion: having complex tube/loop routing will not net you significant better temps... but if it makes you feel better, please do go ahead
Do be aware that the more complex the loop, the more potential failure points exist.
Also, if you have to rely on complex routing to drag that last 0.5°c out of it.. you are doing it wrong
It's like the queeste for the "best TIM" which will keep my CPU 0.5°C under melting point.
Popquiz (for everybody):
Starting condition:
Ambient temp 25°C
Radiator: Copper tubes/copper fins , 120mm , 60mm thick, 11 FPI
Fans: Corsair SP120 in push/pull at full speed
Watertemp: 50°C
How "hot" will the air be when it exits the radiator?
20°C , 25°C , 30°C , 50°C , 100°C , or another number?
.