Hmmm, how do you guys explain the massive amount of heat that pours out of the radiator when you're pushing the system vs not then? Obviously there's a considerable delta in the hot vs cold water? Or are you saying it's not actually as the full body of the water heats in such scenarios and the traversal through the loop/rad only drop it somewhat? I'm actually more shocked that there are no animations or analysis of this jet plane design if its so revolutionary. You have two dies contacting the cold plate pouring localized heat into them with some differential, the problem I'm trying to solve is how these inlet/outlet designs ensure inlet water goes to each die equally and out in a region or fashion that collectively grabs the hot water from both equally? Can this be done w/ any contact block design. Do you guys (with custom water blocks) see temperature differentials between dies when you're pushing your setups based on where the inlet/outlets are (cold->hot) side? It just seems moreso that this mechanism was meant for one central die : Inlet cold water on it and outlet it off die. With threadripper, it appears they didn't or couldn't resolve this fundamental ideology for two spread dies and thus just scaled everything. Problem is : You have inlet water pouring on one die whose heated water then traverse to the other and then captures some more heat and then out? Is this mitigated by water's capability to absorb heat efficiently whereby it still can just as efficiently absorb die #2's heat even in a heated state as good as it did die #1?
The water needs thought of as surface area. You use the surface area of the fins to increase the contact surface of the water, thereby helping transfer heat from one medium to the other. The water is in contact with the surface for fractions of a second before moving on. During that time, heat is transferred to the water. The newly heated water then travels to where it makes contact with the tubes in the radiator, transferring the heat from the water to the tubes. Those tubes transmit heat to the fin array where it sits until air moves past the fins, the contact transferring heat to the moving gas.
This is why when adding radiators, all you are doing is lowering the delta between the air and the water temps. You can only get that to a certain point, beyond which you get little gain. Yes, it adds liquid to the system, thereby adding surface contact to the heat plate, but the water will only pick up so much in that time. This is why they discuss equilibrium after 30 minutes and that, because it can only pick up and transfer out so much, that at some point you cannot do any more cooling with more radiators.
Now, when systems had higher flow restrictions and lower flow rates, the water was in contact for longer periods, resulting in higher water deltas between the inlet and outlet. That is when it had more impact. Modern systems with 1-2 gallon per minute flow rates and low flow restriction blocks change the equation.