Thermal PerformanceTesting a block for thermal performance is fairly simple once you realize that you have to measure VRM temperatures manually. As such, I installed an Omega NTC type thermistor on VRM 1 and connected it to an external display for a VRM temperature readout. TechPowerUp GPU-Z was used to monitor GPU core temperatures. The GPU was overclocked to 2 GHz, although with how GPU Boost 3.0 works, it did vary by +/- 1 clock bin (13 MHz). Similarly, with core voltage being nearly impossible to set manually and fix at that point, it is best to compare the results below within the data set and not to other reviews elsewhere. For what it is worth, 1.20 V was set at Vcore using EVGA Precision-X (and after a shunt mod).
Everything required was placed inside a hotbox, and the ambient temperature was set to 25 °C. Gelid GC-Extreme was used as the thermal paste of choice because not every block comes with TIM included, and cure time was taken into consideration. Five separate mounts/runs were done for statistical accuracy and to remove the chance of any mounting-related anomalies. For each run, a 60 minute Unigine Heaven 4.0 run was done, and temperatures were monitored until a steady state was reached, after which they were recorded. A delta T of GPU core/VRM and loop temperatures was thus calculated for each run, and the average delta T that was then obtained across all five runs. This way, the cooling solution is taken out of the picture.
Note again, in case you skipped the previous page, that I have separate entries for blocks by themselves and with their respective backplates, as sold separately. Those without a backplate entry specifically mentioned either do not have one available, have one included with the block itself or, in the case of the EVGA Hydrocopper, are installed with the stock backplate itself.
When the difference between all the blocks in terms of GPU core cooling is just 3 °C, you know things are hard to distinguish. The CORSAIR XG7 GPU block does really well. I will admit that I had some doubts when I first saw the block over a year ago, and then again when my sample arrived. But, at least on the GeForce GTX 1080, it did really well. Keep in mind also that the backplate does not have any thermal pads here, so it does not actively participate in cooling, and the GPU core is cooled pretty much by the block itself. The lack of a participating backplate may be why VRM cooling was not as relatively impressive, however. While VRM cooling below a certain number does not do much for the GTX 1080, look at these results more of an indicator of where things would be for other cards where VRM cooling helps more. Regardless, everything is well within spec, as expected, and I managed to even sneak in some gaming and benchmarking time while the system was running to make sure nothing was off.