My question would be why is the radiator mounted upside down?
The same (i think) is true for the CPU. If you mount it this way around chances are high that air gets trapped in the CPU block.
Air won't get trapped in the CPU block as long as it's not the highest point in the system. The channels in the block are relatively small compared to the flow rate so bubbles are forcibly evicted very quickly after turning on the system.
In this particular instance, I am going to guess that
@Darksaber was limited by the AOI in question, something that he couldn't change as the standardised test bench requires the exact same hardware to be used every single time.
- There's a drive/cable shroud that comes with the case which obstructs the radiator inlet/outlet if flipped so that the tubes are at the bottom. Since this is a case test and not an AIO test, all parts of the case should be used and the shroud had to stay.
- This particular AIO has a pump in the tubes near the radiator inlet/outlet, not on the CPU block like most designs. That means that with the CPU block rotated it would almost certainly have the pump as the highest point in the loop which is VERY BAD. with the CPU block connections at the bottom, it pulls the cables down, and stops the pump from being the highest point in the loop.
And yes, for your own system, pump at the bottom, sealed header tank end of the radiator at the top is the best/quietest/longest-term choice but for a test bench that's going to be dismantled in a couple of days, I don't think you need to worry about it; It's not going to have caused any meaningful temperature/noise differences that would show up in the test results.
