Skip to main content

33 CPU cooler group test preview: cool, quiet and both

When you buy a processor it almost always comes with a CPU cooler these days, with as a result that the affordable CPU cooler market has pretty much disappeared. The market for high-end and fancy coolers is thriving, however, since more powerful coolers are needed when you want to overclock a CPU.

Noise is another reason to not use the stock cooler. While they usually aren't extremely loud, separately sold coolers are often more silent. The combination of good cooling performance with silence is what makes a good CPU cooler.

We now employ professional CPU simulators that replicate the exact heat production and signature of a Socket 1155 Ivy Bridge and Socket 2011 Sandy Bridge-E processors. They also fit into the official processor sockets and produce heat the same way a real CPU does. That means that the Socket 1155 has separate heat generators for the CPU and GPU parts. The temperature sensor is now integrated into the CPU simulator. The temperatures that we record are therefore representative of the heat produced by real processors.

With our Socket 1155 Ivy Bridge CPU simulator we tested CPU coolers with a various amounts of heat: 65 W, 95 W and 125 W. The first two are the TDP values of the majority of Socket 1155 processors. 125 W equals about what a reasonably overclocked processor will produce. With the Socket 2011 Sandy Bridge-E CPU simulator we test at 130 W and 160 W. The first is the TDP of the actual Socket 2011 processors, and the second corresponds to an overclocked version. We perform each measurement twice, once with the CPU cooler fan running at full speed (12V) and once with the fan running on low speed (7V). We normalize the results to an ambient temperature of 20 degrees Celsius. We also measure how much noise a CPU cooler creates with a professional Brüel&Kjaer 2238 sound level meter.

It's striking how little innovation there has been in the area of CPU coolers during the past two years, with one exception. Cooler Master used a so-called Vertical Vapor Chamber in its TPC-800 cooler, a type of large, flat heatpipe. It works the same as a traditional heatpipe, where the heat of the CPU warms up the liquid inside, which then evaporates and transports the heat to the heatsink. When it cools it runs back down the heatpipe or vapor champer.

The heatpipe remains the tried-and-true method for cooling CPUs. The coolers in this test have two or more heatpipes, with the Silentmaxx BigBlock containing a record number of 10 heatpipes. Thicker 8mm heatpipes are becoming more common, and they are able transport more heat than the traditional 6 mm heatpipes. Their drawback is that you can't fit as many of them.

The amount of copper and aluminium affects how well a cooler can cool. Basically, the larger the cooler, the more surface area it has, and the better it can get rid of heat. It's no surprise then that the best coolers in this test are huge. The design and the types of fans that are used also have an influence on cooling performance.

Most of the coolers in the round-up are tower coolers, a square or rectangular cooling block that is connected via heatpipes to the baseplates. One or two fans blow air through the heatsink towards the back of the computer. The other type, top-flow coolers, blow the air down towards the processor, but these are not very common.

The 33 coolers in this group test come from 13 different manufacturers, and we divided the test results into three different categories based on pricing. You can find the full test on Hardware.Info.