Performance
The performance of the NH-L9i-17xx coolers with and without the NA-FD1 fan duct has been tested with the i5-12600K processor, ASUS Strix Z690-i Gaming WiFi motherboard, Kingston Fury Beast 32GB DDR5-5200 memory kit, and Patriot P4 M.2 NVMe SSD. A PC case used for tests is more important – Jonsbo A10 in a version with perforated side panels.
There are three settings: idle, which explains itself, a mixed load test, which includes a pretty long PCMark 10 Extended benchmark, and a max load CPU test with AVX/AVX2 instructions, performed by the AIDA64 stability test.
The results require additional explanation. All tests were performed at automatic motherboard settings. In theory, the i5-12600K runs at a 125W with maximum turbo wattage of 150W. In reality, the maximum wattage during work differs depending on the motherboard and additional variables. In our tests, the maximum temperature was limiting the maximum frequency during AVX/AVX2 tests, but in games or tests that simulate daily work, the CPU could run at its declared performance, up to 4.9GHz.
The CPU was going up to 88W during PCMark 10 tests and could keep 4.9GHz, while during maximum load tests, the maximum frequency was going down to 4.5GHz at about 115W.
I’m surprised how well the NH-L9i handles the i5-12600K CPU. We can still see some throttling in AVX/AVX2 tests but let’s be honest; barely anyone uses these CPU extensions. The CPU itself can make 300-400MHz more with the NH-L9i than the last generation, which is a significant improvement.
Both Noctua NH-L9i-17xx coolers were performing the same, so you can see it as one result in the comparison. However, I’m sure that some readers would be interested in how the NA-FD1 duct performs in the push and pull scenario. The standard option, optimized with the included mounting kit, is a push. In tests, it performs slightly better, up to 2°C. In some specific PC cases, you may consider using pull orientation as it may help to cool other components on the motherboard. It wasn’t really visible in our case, but the used Jonsbo A10 case also offers good airflow.
The final results are better than expected. Noctua says that the difference because of the NA-FD1 air duct is about 5°C. In our tests, we could see a 6-7°C difference, which is quite a lot considering there is nothing else used. At the given price of $13, it’s for sure worth to equip our small PC with the NA-FD1. It’s also surprising that something as simple isn’t manufactured by anyone else. I remember old ATX cases with built-in air ducts as a recommended addition to Intel Pentium 4 processors. Some years later, this idea died, but it can still help in some specific scenarios like more popular recently small form factor PCs.
