We explored the best fan and radiator setups for the Corsair Air 5400

Without further delay, let’s get into the hardware specifications that were used to run the tests, and the testing methodology:

CPU cooler: NL-LC1-36 + NF-A12x25 G2 PWM chromax.black
Supporting case fans: 6x NF-A12x25 G2 PWM chromax.black
CPU: AMD Ryzen 7 9800X3D (with offset mounting)
GPU: GIGABYTE AORUS GeForce RTX 5090 Stealth ICE 32G
Motherboard: GIGABYTE X870 AORUS Stealth ICE
RAM: Crucial Pro Overclocking White UDIMM DDR5-6000, CL36
SSD: Lexar NM1090 Pro 1TB
PSU: Seasonic Focus ATX 3.0 GX-1000W
Case: Corsair Air 5400 White

The main goal was to compare different AIO installation scenarios in the dedicated chamber and evaluate any potential performance or noise implications. We performed multiple tests, but we will focus on the scenario where the GPU fan speed was fixed at 1300 RPM, which was deemed the sweet spot for cooling efficiency relative to noise, and the pump speed was fixed at 50% PWM, while all fans were cycled at 25%, 50%, 75%, and 100% PWM speeds.

It’s important to note that running the NL-LC1 pump at a fixed speed is not something we generally recommend, a topic we cover in our FAQ section, but in this case, it was helpful to ensure fully stable test conditions. Stress testing was performed using OCCT with a combined CPU and GPU load for 30 minutes per scenario, and each run was preceded by a 60-minute warm-up cycle.

All temperatures were captured in a controlled environment as a delta over the ambient baseline temperature, which averaged 21.8°C throughout the testing.

As noise emissions were one of the key focuses of this test, measurements were taken starting at the front of the case and moving around it in 45° increments, and the averages are presented. That makes 8 noise measurements, just for one scenario at a given fan speed.

To set the baseline of the test, we ran push and pull fan setups with the radiator mounted in the dedicated third chamber to see if there would be any CPU temperature differences. When using fans in pull mode, it can often be beneficial to use our inlet spacers, which move the fan 5mm further away from the radiator fins, so we tested a setup with the NA-IS1-12 installed as well.

Using the NA-IS1-12 spacers can help to reduce undesired influx turbulence in suction-type applications where the fan draws air through an obstacle such as a radiator. This reduction in turbulence can have significant positive effects on both acoustics and performance, so it was interesting to see if that would improve acoustics in this case, pun intended.

Looking at the CPU temperatures, the fans in the pull setup outperformed the push setup by an average of 3.2 degrees across the entire fan speed range.

While this difference didn’t show any real performance or clock speed benefits in our testing, it’s definitely worth pointing out, since this result is quite different from what we’ve seen in our tests with other cases. We usually measure similar temperatures between push and pull, but with a clear acoustic advantage for push, which is why we generally recommend a push setup on the NL-LC1 series liquid coolers. The third chamber of the Air 5400, in combination with our fans, clearly behaves differently in this regard.

As with all temperature advantages, the performance headroom in this case can be traded for lower fan speeds and less noise, putting the pull setup in a favoured position.

The use of the inlet spacer becomes relevant when we look at the noise measurement data. While having the fans in pull mode behind the grill, filter, and radiator provides a clear temperature advantage, adding inlet spacers to the pull setup brings a slight acoustical advantage.

If we take a closer look, the average noise reduction advantage of the inlet spacer hovers around just 0.5 dB(A). But if we go back to the CPU temperature measurements, we can see that it came with an average 2.3°C penalty across the board, which for many will not justify the investment in the spacers. While we are only measuring pure sound pressure levels, even subjectively, installing the inlet spacers didn’t noticeably improve the perceived sound profile either.