Originally Posted by prava
Fans matter a lot, specially when the difference in rpm is so huge. The pump barely matters at all, specially in such simple loops. Should you try to run the H100i pump with a full loop you would fail, but for what its used is powerful enough.
Don't get me wrong, I hate loud systems and thus why I'd take the H220 if I ever go looking for an AIO solution.
You are using improper wordings. Push or pull makes no reference to whether you use your rad as an intake or exhaust, but whether you are "pulling" our "pushing" air with the fan.... and by the looks of the video, they were actually PUSHING air out of the case, not pulling it.
And no, the difference is ACTUALLY HUGE if you plan to use the rad as an exhaust instead of as an intake because the air temperature difference will be abysmal... and you would have to plan your layout carefully because you would probably starve your rad for air... to the point you might have to turn the back fan into an intake.
It is a terrible idea, and laws of physics don't apply when you have a fan pushing air, it simply doesn't matter that hot air rises, because the temperature difference would have to be HUGE to go against a fan or make it any slower.
Also, you are missing the point entirely: the hardest part of your system to get cooled is, by far, the cpu. Ram temperature has been meaningless since we started using DDR2 (and you have direct flow on them so, actually, temperatures would be OUTSTANDING here), and cooling your gpu will be really easy in the simplest of the case if you have a fan pointing at it.
Again, you are totally wrong, and you clearly have no idea of what you are talking about. Let me clarify:
a) That 1ºC difference is measured at the rad.
b) The heat is calculated based on a heater and how much power its drawing.
c) The tools used for measurement are precise enough to grant far more precision than a simple 1ºC
d) Even if we were talking about a cpu, 1ºC is clearly NOT the margin of error, not when you have the proper tooling. If you average the temperatures of your cpu for a big enough period of time you can find out where it fares once you average them. Keep in mind that cpu sensors are very precise although they aren't accurate, but for this matter this is simply perfect.
I'd suggest you get your facts straight and together before engaging in things you don't totally understand...