To make things simpler, I have decided to use a CORSAIR Hydro XD5 pump/reservoir combo unit rather than a discrete pump and reservoir. The pump is powered by a direct SATA connection to a CORSAIR HX750 PSU and controlled by an Aquacomputer Aquaero 6 XT. There is a previously calibrated in-line flow meter and Dwyer 490 Series 1 wet-wet manometer to measure the pressure drop of the component being tested—in this case that of each radiator. Every component is connected to the manometer by the way of 13/16 mm tubing, compression fittings, and two T-fittings.
As of the time of this review, I have tested seven entries in total. Unless specifically mentioned, every one was tested with an accompanying backplate. As that was the majority of them as of round one, this will not change in the future, either. Those without a backplate specifically named in the plots come with one included with the block and not as an optional purchase. All seven have now had dedicated reviews, and more are being tested right now. So keep in mind that the highlighted entry is for the Watercool Heatkiller GPU block and backplate combination.
Of course, I would say that a backplate generally does not matter as far as liquid-flow restriction goes, but this generation has been nuts. Not only have we had some very interesting water blocks come out for the NVIDIA Founders Edition cards, but the hot VRAM on the back of the RTX 3090 PCB has caused enough interest for a couple of companies to offer active backplates for the first time. This is why there is a separate entry where an active backplate is added.
I do not expect much to change with the copper cold plate, though there may perhaps be a slightly lower pressure drop owing to the nickel-plating no longer taking up some room in the microchannels. As such, the results here will be indicative of the entire Heatkiller V range for the RTX 3080/3090 reference PCB blocks, and we see it is slightly more restrictive than average owing to the more complex cooling engine and higher number of fins.