I just wanna point out, when he says 5C he almost certainly means in GamersNexus testing.
That's in the context that 2C from the first post here isn't applicable because the testing methodologies are insufficiently refined, but that GamersNexus metodologies are in fact good enough and close to ThermalBench and the 2 websites Doyll mentioned.
I also want to point that GamersNexus is still technically flawed, and doesn't represent the best case scenario (For Air coolers especially) in an actual case in an actually completely and fully airflow optimized PC.
My opinion that the difference is, in fact, closer to 2-3C (or even non-existant) when you have the best mesh-metal vented case, the best-performing fans in front, the appropriate amount of fans, best placement and optimization (including 25mm thick fans on the GPU without surround directly on the heatsink) and the most perfect custom fan curves for your setup.
Also including discretely taping out the intake parts of the case (to the extend that satisfies personal aesthetic needs) as Doyll always recommends, and having 1-3/or all PCI-E brackets removed around the GPU, and or a custom filter placed there.
To get NH-D15 and similar to match best CLCs takes more work, more planning, more thinking, but is absolutely possible.
Last edited by Dogzilla07; 06-22-2020 at 05:19 AM.