{"id":227080,"date":"2025-03-19T21:00:37","date_gmt":"2025-03-19T20:00:37","guid":{"rendered":"https:\/\/www.hwcooling.net\/?p=227080\/"},"modified":"2025-03-20T01:16:45","modified_gmt":"2025-03-20T00:16:45","slug":"more-than-100-models-tested-overview-of-m-2-ssd-coolers-review","status":"publish","type":"post","link":"https:\/\/www.hwcooling.net\/en\/more-than-100-models-tested-overview-of-m-2-ssd-coolers-review\/","title":{"rendered":"More than 100 models tested. Overview of (M.2) SSD coolers"},"content":{"rendered":"

It\u2019s time for a comprehensive recap of the results from SSD cooler tests we\u2019ve accumulated over the years. We\u2019ve gathered quite a lot of data, with most of the coolers being models integrated into motherboards, compared to standalone designs available for purchase. The following article primarily creates a database of various SSD coolers, based on which you\u2019ll be able to determine how each available model performs in terms of cooling efficiency.<\/p>\n

In-depth motherboard analyses<\/a> are, among other things, a good source of information about cooler efficiency. Compared to third-party models from Akasa (Gecko Pro<\/a>), Arctic (M2<\/a>), Axagon (CLR-M2Lx<\/a>, CLR-M2XT<\/a>, CLR-M2<\/a>, CLR-M2XL<\/a>), BeQuiet! (MC1<\/a> and MC1 Pro<\/a>), EKWB (EK-M.2<\/a>), Gelid (IceCap<\/a>), Savio (AK-60<\/a>), and Thermalright (HR-10 2280<\/a>), you can get an idea of how each cooler reduces SSD temperatures.<\/p>\n

Certainly, we test only with one SSD under unique conditions, and depending on these variables, the results may vary slightly. However, you can still get a rough idea of the efficiency of individual coolers. More powerful models will dominate even with higher airflow in the system, while weaker ones will lag behind.<\/p>\n

\"\"<\/a><\/p>\n

When evaluating typically monolithic designs, it is often assumed that a larger surface area means higher cooling performance. However, this is not always the case, as there are variables such as the thermal conductivity of the thermal interface material (TIM) and mounting pressure.<\/p>\n

Mounting pressure can sometimes be very good, with a high heat transfer coefficient, but it can also be weaker. This is usually due to only partial contact with the SSD, which is typical for technically underdeveloped tool-less systems. While you can install the cooler conveniently, without the need for a screwdriver, its contact with critical components (typically the SSD controller) will be weaker.<\/p>\n

The above may not always apply, of course. It also depends on the design of the SSD itself\u2014thicker models (SSDs) may have better contact (i.e., contact over a larger area) with the same cooler than thinner ones. These things can vary from case to case, and it is up to each manufacturer to address the situation.<\/p>\n

\"\"<\/a><\/p>\n

Our tests are based on measurements taken on a single SSD. The testing methodology is described in more detail in the subsection below.<\/p>\n

Methodology<\/h2>\n

The tests are conducted in a wind tunnel, which replaces a standard PC case<\/a>. For system cooling, four Noctua NF-S12A PWM fans @550 RPM are used in a balanced ratio of two intake to two exhaust fans. The intake air temperature is strictly controlled and, for maximum accuracy, is maintained within a narrow range of 21\u201321.3 \u00b0C.<\/a><\/p>\n

In accordance with findings from measurements on the impact of different positions on cooling efficiency<\/a>, standalone SSD coolers are tested in the first slot, above the graphics card. From motherboard tests, we naturally have many results for coolers (labeled as cooler 1, cooler 2, etc.) in other positions, but these are determined by their fixed locations.<\/p>\n

\"\"<\/a><\/p>\n

The SSD without a cooler is tested on the Asus ROG Strix Z790-E Gaming WiFi<\/a> motherboard in two locations\u2014in “slot 1” above the graphics card and in “slot 2” below the graphics card. From thematic tests where we examined the dependence of SSD cooling on the chosen position, we know that there are certain differences in cooling<\/a>, although it cannot be generalized which is the “more suitable” option. This can always vary depending on factors such as whether the graphics card cooling is active or not (and, of course, on the case fan layout and the overall system cooling characteristics). In our case, with always passive graphics card cooling, placing the SSD in the second M.2 slot proved to be the worse choice in terms of cooling. Due to excessive temperature, the maximum performance was not even achieved, as sequential read speeds on the Intel platform in the PCIe 4.0 \u00d74 slot should typically range between approximately 6600\u20136900 MB\/s<\/a>. In this case, it was only around 5500 MB\/s.<\/p>\n

\"\"<\/a>
Illustration of SSD placement in the M.2 slot below the graphics card<\/figcaption><\/figure>\n

For testing, we use the Samsung 980 Pro (1 TB) SSD. The load is applied for 10 minutes (which is sufficient time for temperatures to stabilize in the wind tunnel) using CrystalDiskMark\u2014cycles of sequential read and write operations. The power consumption during this time is approximately 6 W, which is the upper limit for M.2 SSDs and should not change even for models supporting PCIe 5.0.<\/p>\n

For completeness, it should be noted that the tested SSD is always in its original state with the label intact. Removing the label is rather unnecessary, as it has little impact on heat transfer to the cooler. We have also covered this in detail in previous tests<\/a>:<\/p>\n