Review on 🌙 Noctua NH-D15 SE-AM4 Review: Premium AMD AM4 Dual-Tower CPU Cooler (Brown) by Shane Ryder

I'll try to be brief. I've worked with laptops for 6 years because I had a need for mobility that trumped all other concerns. Lately that requirement has disappeared, so I've decided to build the best possible PC within reasonable bounds. The AIO piqued my curiosity and thought the NZXT Kraken x62 would be the best cooling solution for my 2700x. This was problematic for a number of reasons, the biggest of which was the noise profile. The sound of the pump annoyed me. Now, the computer was so quiet that when I was at full load, you could interrupt the sound of the pump by scratching your hand, but it was very annoying, especially when I wasn't playing games and the graphics card fan wasn't working. In short, missing the 2700x manual overclock to 4.3GHz for all cores, which HWinfo reported as 230W CPU and SoC consumption, the NH-D15 SE-AM4 was x62 capable. With Performance Enhancer Level 2 enabled, I achieved the same clock rates as the x62. The only differences are: The NH-D15 is so quiet that I have to physically put my ear to the case to hear it and its spare NF-a12x25 fans under full load (small Prime 95 FFT, Folding@home CPU -work units). ) or lightly loaded (games like FFXV, web surfing, etc.) where fan noise has always been present on NZXT. In the worst case, the processor runs at a maximum of 68c when booting with PE 2 versus 64c at x62. Both configurations offer an all-core clock frequency of 4.190 GHz. I have never been able to find a fan profile with stock NZXT Aer P fans or a replacement (as pictured and MUCH quieter than AER P) NF-a14 push-pull, push-pull or pull fan that does the Temperatures below keeps 70°C and is silent. fully loaded. NH-D15 is simply silent with NF-A12x25 fans for all scenarios. Under light load conditions, the D15 is slower and the 4.34GHz multi-core but not the 4.34GHz all-core autoclock is much less likely to drop. The overall performance of a computer under light usage would often be indefinitely faster in various situations than x62, which drops peak performance to the 4.2 range within 5-10 seconds of use and automatic overclocking. When overclocking the NH-D15s manually, my particular example of the 2700x is (barely) behind the AIO in absolute performance. I suspect that the surface area of the CPU die under the heatspreader and the presence of heatpipes just above that contact pad is the main reason for the difference, but that's speculation. An observable and repeatable fact is that the D15 runs significantly hotter than the x62 when the CPU + SoC wattload is above 150W. As the wattage increased, so did the gap between x62 and D15. It's not too fresh in memory, but the 4.2GHz iirc sees 170W+ drained and an average of 70s C on the D15, where the Kraken x62 could hit the lofty 60s. At 4.3GHz in manual mode and a whopping 230W (HWinfo+ a few weeks after test runs) the 2700x remained 100% stable but was 84-87C, while x62 could reach 75 but was mostly around 70C . @ never stable for me 4.35ghz and consuming 265watts before freezing x62 will never get close to 80c when system shut down hard and shut down with nh-d15. Overall, I prefer to run with quietness and power consumption that will outlast aging to a "top speed with a little longer death run" configuration and find the NH-D15 the clear winner in all aspects of performance. This is of course with the exception of people who are likely to need to use closed circuit chillers. Pure silence and all that performance at an undetermined 4.34 multi-core clock speed when surfing the web beats x62 pump noise, rapid frequency throttling under light loads, and a very, very small specter of a one-day leak. On top of that, the NH-D15 SE-AM4 was much easier to install than the AIO. I wonder why bother with anything else?