Best Laptop Cooling Pads Tested


We tested six of the most popular laptop cooling fans to get an answer to the burning question, “how well do these things actually work?” The Cooling Fans in this test are:

Coolers that are capable of more than one fan speed were tested at both their lowest and highest settings. Coolers with only one speed appear in both low and high graphs.

Acer Testbed

Lenovo Testbed


The hands down winner of this competition was the SF19, delivering at least twice the cooling factor at its highest output, and competitive even at its lowest output. This is proabably mostly due to its superior electrical power source. The SF19 drew 12W of power at its highest setting, four times that of the next most power hungry model, the three-fan Nexxtech / Havit HV-F2056 / Avantek CP173. However due to the fact that it must be plugged into mains power to work, it seems that the SF19 has enjoyed less popularity than the other models which are powered over their usb bus and as a result the SF19 has become very hard to find lately. When you consider that the most common use case for an external cooling pad is probably during gaming or other cpu intensive operations, and considereing that a laptop would likely be plugged into mains power then anyway, we believe that the lack of usb power is not the drawback that many might suppose it to be. If you find one snap it up. Otherwise if you want the best cooling power available we might have to wait for usb “C” models to arrive, which are the logical next step and will be capable of delivering sufficient power over the usb cable alone.

The Nexxtech model is identical to two other popular models, the Havit HV-F2056 and the Avantek CP173. The build quality is decent with a metal top mesh and sturdy construction. It has an on/off switch on the back along with an extra usb port, and blue led lighting. Be aware though that the led lights cannot be disabled and are always on while the unit is powered.

The Targus is competitive for its efficient use of power and is also very quiet compared to anything else in the running (i.e. everything other than the CMC3). It is an all plastic model designed to be held on the lap comforably and to that end it is complemented by a soft gabardine pad on the bottom.

Acer Dirty Testbed

At the beginning of the experiment the condition of the Acer laptop presented us an interesting situation. Its cooling system had not been serviced in some time and had accumulated some dust both on the fan and inside the grill. This allowed us to run the tests on the Acer twice, first in this dirty state, and then again after cleaning. The results are illuminating as you can see above. What surprised us most is that the external cooling pads were actually more effective on the dirty state! Of course the temperatures were much higher in the dirty state, but the external fans made a much larger difference as well. Note also that in order to run stress-ng without overheating the laptop in its dirty state, the software had to be configured to stress the cpu at a reduced rate. The “dirty tests” were run at 70% work load. All other tests were run at 100% load. Note that the Nexxtech cooler was not tested in the Dirty state.


We measured the sound output of each Cooler at a distance of 0.5 meters. Results are in below histogram. The sound floor of the sound meter we used is 30 dBA, and the ambient noise in the test environment (quiet basement at night) was 33.3 dBA, so the graph begins at that ambient level.

Power Draw

We also measured the power consumption of each model using the USB Detector, except for the SF19 which was measured using the EM100 energy meter. Results are below.

Test Methodology

Two different laptops were used as testbeds for this experiment. The tests were carried out in cycles of work and then cooling. This was accomplished by loading the laptops’ CPUs using stress-ng and running them in this loaded state while repeatedly measuring the laptops’ built-in temperatures sensors’ output (once per second) until the 800 second moving average temperature became equal to the 400 second moving average temperature. At that point stress-ng was terminated and the laptop was allowed to cool until again, the 800 second moving average crossed over the 400 second moving average (this time in the opposite direction). All tests were performed on an identical Knoppix Live CD image. The laptops chosen for this test are the Lenovo x201 with an Intel i5 processor, and an older Acer Aspire 1694WLMi with an Intel Pentium M processor. Both laptops had their hard drives removed before the tests were performed.

All tests were carried out in a temperature stable environment where the ambient temperature varied by a maximum of two degrees over all tests. All temperature were then normalized for any differences.