Originally Posted by
churchy
I somehow didn't get how rad length got converted to head dissipation capabilities. sum in millimeters of one rad length (not area, length) + another rad length + length of third rad (forgetting it has less width) somehow got equal to same number but in watts (ignoring rad model cooling performance, no specified temp deltas, fan speed, ammount/way of setup of fans, different performance of 140 vs 120mm fans), then sum of cooled components heatdump got divided by rad length?
OK, i get what you wanted, so this dissecting of calculations was more meant as harsh joke :), but if you wanted to divide rads proportionally to cooled components heat dump you should at least (if rads are of same model line) count their area. as in instead of 560 use 4*140^2, 1*120^2 and such. And to find out proportion one has no need to replace mm2 with W :)
+ascl already noted, that cpu o/c is more low temp needy then gpus, even though heatdump is similar. Though then again, with that big rad area i doubt both ways will differ much more then 2-4 degrees C be it of single or dual loop. Also nothing beats real world testing numbers - you can simply try both setups.
