PC hardware modder Trash Bench has demonstrated an unconventional gaming laptop cooling experiment using a blower designed to inflate bounce houses, reducing GPU temperatures from approximately 84°C to 48.8°C. The test involved a laptop equipped with an Intel Core i7-10750H processor and Nvidia GeForce RTX 2070 Max-Q GPU, with Cyberpunk 2077 performance increasing from 47 FPS to 56 FPS across the cooling comparisons.
The experiment, documented in Trash Bench’s YouTube video titled “How Far Can You Push Laptop Cooling?”, compared a commercial Llano V12 cooling pad against several custom fan configurations. Rather than modifying the laptop’s internal cooling hardware, the modder focused entirely on forcing additional air through its existing ventilation system.
The results demonstrated substantial temperature reductions, although increasingly powerful fans produced progressively smaller gaming performance improvements.

Trash Bench began with the Llano V12, a commercial cooling stand that uses a fan and foam sealing system to direct airflow into a laptop’s underside vents. At medium fan settings, the cooler reduced average GPU temperatures to 58.3°C, compared with approximately 84°C without external cooling.
That represented a temperature reduction of roughly 27°C, accompanied by an average gaming performance improvement of 19.4% in the modder’s tests.
Increasing the Llano V12 to maximum speed reduced GPU temperatures further to 56.7°C. However, the additional airflow generated considerably more noise without delivering a meaningful increase in gaming performance.
The next configuration used a Noctua NF-A20 200mm fan connected to a custom 3D-printed duct. Foam around the duct helped create a seal against the laptop, limiting airflow escaping before reaching the intake vents.
Despite the fan’s large diameter, the GPU averaged 66.3°C, making the setup less effective than the commercial Llano cooler.
Trash Bench then tested six Arctic P12 fans operating at full speed. This arrangement lowered GPU temperatures to 57.4°C, slightly outperforming the Llano V12 at medium speed but falling short of its maximum-speed result.
A subsequent configuration using two 4,000 RPM server fans delivered another improvement, reducing temperatures by an additional 2.6°C compared with the Llano V12 operating at maximum speed. The server fans were also substantially louder than the previous arrangements.
For the final test, Trash Bench connected an industrial bounce house blower to the cooling stand using a flexible duct and a custom 3D-printed adapter.
The blower generated enough air pressure to physically lift the laptop away from its cooling stand. The modder had to tape the laptop down to maintain a seal and prevent air from escaping.
With the blower operating, average GPU temperatures reached 48.8°C, approximately 35.2°C below the original 84°C reading.
Despite the substantial thermal improvement, gaming and benchmark performance showed diminishing returns.
In 3DMark Time Spy, the bounce house blower, six-fan Arctic configuration, dual server fans and maximum-speed Llano V12 produced results within approximately 1% of one another. The Noctua setup trailed the blower configuration by around 3%, although it operated much more quietly.
The comparison illustrates how cooling can improve gaming performance when a laptop is limited by temperature, but further reductions do not necessarily translate into higher frame rates once other constraints become more significant.
The experiment also highlights the difference between lowering component temperatures and improving overall gaming performance. The industrial blower delivered the lowest GPU temperature, but its performance advantage over substantially smaller cooling systems was minimal.
Extreme cooling experiments are not limited to laptops. Another PC enthusiast used an outdoor radiator to cool an RTX 4070 Ti Super, demonstrating a similarly unconventional approach to reducing gaming hardware temperatures.
Trash Bench’s results are specific to the tested laptop and cooling configurations rather than a universal measure of external laptop cooling performance. They nevertheless show that a conventional cooling pad can capture much of the available performance benefit without requiring the noise, pressure or impractical hardware associated with an industrial blower.
