PC cooling experiments are strange and plentiful. From Peltier modules to giant bellows, liquid nitrogen, tap water, and direct immersion, you might think we've tried everything—but there's always room for more ideas. That's how we ended up at the German forum ComputerBase, where user AssassinWarlord published a long list of results obtained with different interfaces... including cheese, mashed potatoes, and ketchup.

Every computer needs to dissipate heat somehow. Little or much, it doesn't matter: the goal is to reduce the temperature of its components, and we have several strategies at our disposal. The simplest is to do nothing and let a chip dissipate heat on its own—common in low-power systems—but in most cases we use fans and metals like copper, aluminum, or a combination of both to improve the process.

The problem is that's not enough: the contact between the chip and the metal must be optimal, so we need a thermal interface. Over the years we've evaluated all kinds of substances, always associated with magical cooling promises, but if you want to see the latest round of experiments, you'll need to visit the German forum ComputerBase. There, user AssassinWarlord published results for over a dozen and a half interfaces, based on a Radeon R7 240 graphics card with a 30-watt TDP. First conclusions? Ketchup isn't that bad.

Cooling a Graphics Card with Ketchup and Other 'Thermal Pastes'

Ketchup as Thermal Paste and Other Experiments: How Hard Is It to Cool a GPU?
Ketchup as thermal paste
Ketchup as Thermal Paste and Other Experiments: How Hard Is It to Cool a GPU?
The contact of ketchup with the aluminum block (AssassinWarlord)

Among the 'professional' materials we see various thermal pads with different thicknesses—Arctic, Alphacool, and EKWB are some of the brands mentioned—but when we move into the DIY space we find ketchup, mashed potato, toothpaste, baby cream, and... cheese.

Ketchup as Thermal Paste and Other Experiments: How Hard Is It to Cool a GPU?
The FurMark test for ketchup (AssassinWarlord)
Ketchup as Thermal Paste and Other Experiments: How Hard Is It to Cool a GPU?
The general chart that 'AssassinWarlord' published with all the results

The best results are the most obvious: Arctic MX-4 and Corsair TM30 pastes appear in the top spots with 49 and 54 degrees Celsius respectively, while the podium is closed by the Arctic TP-3 1mm pad with 61 degrees. However, the most impressive performance among DIY was demonstrated by ketchup, keeping the card under a FurMark session at 71 degrees. That was enough to place ketchup in the Top 10 of the tests.

Ketchup as Thermal Paste and Other Experiments: How Hard Is It to Cool a GPU?
Obviously, a test with toothpaste couldn't be missing. Sensodyne, in this case. (AssassinWarlord)
Ketchup as Thermal Paste and Other Experiments: How Hard Is It to Cool a GPU?
90 degrees under FurMark, 'enough' to avoid throttling (AssassinWarlord)

Only toothpaste managed to escape the throttling condition with 90 degrees. Cheese, potato, and baby cream took the R7-240 straight to 105 degrees. At the same time, high-profile failures draw a lot of attention, like copper paste, the EKWB pad, and copper and aluminum tapes. So, you know: if you have a low-power component and the weekend caught you without thermal paste... you might find a temporary solution in your refrigerator. (?)

Source and images: AssassinWarlord on ComputerBase