Introduction

As technology continues to advance and we observe faster and faster hardware, there comes a time when upgrades to test systems are needed. Thus, we are upgrading our CPU cooler test bench from the current Intel Core i9-10900K and AMD Ryzen 9 3900X-based test systems to all-new ones. New process nodes deliver greater transistor counts and, consequently, greater heat generation in smaller and smaller dies, making the process of keeping them cool more challenging. Meanwhile, in the quest to deliver more performance, it is accepted that overall power draw is increasing, and new technologies like chiplets are introduced. Therefore the new test systems take all of this into account, along with an entirely new way of testing coolers that is detailed below. All these changes have been made in order to provide you with an even more in-depth look at the performance of reviewed coolers and how they behave with various workloads.



With the latest system update, we will continue testing on both AMD and Intel systems. Despite the added complexity and time consumption, testing on both remains, in our view, the appropriate choice for the time being, as both AMD and Intel have taken different directions at the moment. While that will certainly change in the future, we are testing with what is available today. Intel's monolithic processors remain power-hungry at the high end. Meanwhile, AMD's current generation, with its chiplet-based design and thicker IHS, represents a different kind of beast in terms of cooling needs. In this article, both systems will be detailed in-depth along with the new testing methodology.

Testing Systems, Platforms and Settings

The New Test Method

No matter which exact processor model you choose for your rig—entry-level, midrange or high-end, AMD or Intel—they all come with a TDP rating from the vendor. Unfortunately these ratings are not standardized and vary greatly between vendors, and even between product generations. The first novelty with our new cooler review test setup is that we'll be testing each cooler at several TDP levels on both platforms, to provide you with data not only how it performs with the most powerful CPU, but also with something that's used by the more-average gamer.

While all modern processors have mechanisms that report the CPU power consumption, these readings are hardly accurate, and they also depend on various tweaks applied by the motherboard vendor and the BIOS settings (Load-Line Calibration, or LLC, for example). Instead we opted for a hardware measurement solution. This approach allows us to physically monitor the processors' power draw, ensuring a significantly higher degree of accuracy. By directly monitoring the CPU's power usage, and integrating this data into our measurement pipeline, together with fine-grained control of the load parameters, we can provide results based on hardware-confirmed TDPs rather than relying on software estimates.

On a fundamental level, the testing method used will identical for both AMD and Intel. There will no longer be set clocks, voltages, or testing of cooler performance at base and sustained boost clocks. It makes more sense to set a TDP target that is physically monitored during testing. This allows the processor to operate at a controlled power level, allowing us to do performance testing that not only shows how a cooler should typically perform on entry or mid-range processors but also how these coolers will behave in various workloads. The data this provides can be correlated with our processor reviews, which display per-load and average power draw in various tests, including gaming, thus giving users a better grasp on the level of cooling their system needs.

This new testing method, designed to benefit all users, allows for a more practical evaluation of coolers. When utilizing a tower cooler with a 150-watt rating in an overclocking test, failures may occur due to the overclocked CPU generating excessive heat, and it might be dismissed by many as a choice for their setup. However, this scenario does not provide an accurate assessment of the cooler's overall performance, because the cooler was never designed to handle such high heat loads. Consequently, assessing coolers across various TDP settings offers a clearer understanding of their potential. Most importantly, this approach empowers users to choose coolers that precisely meet their needs, avoiding the unnecessary selection of overkill solutions.

Temperature Testing

Idle temperature testing consists of the systems simply sitting idle for some time, and taking the average temperature. After that, we run tests at a specific power target, in 30-watt intervals, starting at 30 W, then 60 W, 90 W, 120 W, 150 W, and so on. We use Y-Cruncher, which generates a consistent and repeatable load on the processor that scales to the highest power levels. After the temperature stops increasing and stabilizes, the average temperature is recorded. This test is conducted three times to ensure consistent results.

Y-Cruncher is used to apply load to the CPUs, with the overall consistency of the workload making it perfect for comparison at each of the designated target TDPs. Once the temperature has stabilized and is no longer showing any further increase, the temperature reading is reset, and the test continues for 120 seconds, recording the average temperature during that time frame. These tests are run three times to check for any issues.

As for the ambient temperature, it is set to 22 °C / 72 °F, which is actively monitored via a 4-Channel K-Type Thermocouple Thermometer to detect any change greater than +/- 1°C in regards to the ambient temperature of the room.

Fan Noise and RPM Readings

During our acoustic testing of CPU coolers, the entire system operates passively—no fans, except those on the CPU cooler, are engaged. The power supply features a semi-passive fan-stop mode, enabling us to concentrate solely on evaluating the acoustic performance of the CPU cooler, as all other sources of noise have been effectively removed.


Note the chart above is for reference only. The coolers shown had their noise levels taken at 6 in. / 15 cm. However, stepping away three feet results in a sizable drop in perceived noise. This was done to accommodate the new testing area, where our noise floor is a bit higher than it was previously. We also expanded noise testing to include 25%, 50%, 75%, and 100% PWM settings. For example, a cooler like the Noctua NH-D15 has a noise profile of 34, 40, 43, and 49 dBA.

It should be noted that different meters will give different results as well. This is why decibel readings between different review sources should not be compared. Plenty of variables are not accounted for—instead of looking at this and previous reviews, it's better to compare where the coolers land in relation to each other to better grasp the noise level you can expect and pick the cooler that is right for you.

Noise normalized testing will remain with all CPU coolers tested at 45 dBA (at 15 cm), which, as noted above, is fairly quiet. At this noise level, a cooler can be considered "pretty silent" by a majority of users, and depending on your chassis of choice. Also note that some coolers do end up quieter than our stated noise normalized settings. However, the majority end up equal to or louder at maximum RPM.

Speaking of RPM, we also record the cooler's fan speeds at the same 25%, 50%, 75%, and 100% PWM settings to give users a direct reference point from which the dBA readings were achieved.