In one of my previous videos a subscriber suggested that I should use mayonnaise as thermal paste. I conducted some research on the subject and it appears that a lot of people have done that already. In an attempt to be somewhat original I sat down, did some brainstorming, and whipped up this crazy idea!
With a tube of “Aquafresh Extreme Clean” in hand I heaved my Gateway 832GM out onto my desk and began setting up for a pretty awesome experiment. The 832GM is equipped with a Pentium 4 Prescott processor running at 3GHz (keeping this monster cool with normal paste is challenging enough), 1GB of DDR RAM, and Intel Integrated 900GMA graphics. For this demonstration I used Windows 7 installed to a flash drive so we could open up a hardware monitor and check out the system temps recorded by the on-board sensors. I also used an IR thermometer to check the temperatures of the heat sink itself.
First, I applied some cheap Banggood thermal compound to the CPU and recorded the system temps. This thermal paste is not the greatest stuff in the world but it gets the job done. While sitting idle the CPU leveled out at 53C (keep in mind this is a Prescott) according to the hardware monitor, and the IR thermometer recorded 40C at the lowest point on the CPU heat sink. After running prime 95 for 10 minutes straight I recorded the temperatures once again and the Pentium 4 had reached 60C according to the monitor and the heat sink was still sitting at 40C.
This next section was the fun part of the video! I removed the heat sink, cleaned all of the thermal compound off, and preceded to squeeze out some toothpaste onto the CPU. By accident I put way too much on the Pentium 4 which made cleanup afterwards a mess. After securing the heat sink back onto the processor I fired the system up and it booted into windows without incident.
Checking the system monitor, I was surprised to find that temperatures were staying close to those that I recorded while using the standard thermal compound. After running the same amount of time at idle the system reported temperatures around 50C and my IR gun clocked in at 37C. Unexpectedly, idle temps while using the toothpaste were slightly better vs. when the thermal compound was being used! The results while the CPU was under 100% load were nearly identical to those that I recorded with the traditional paste. According to the on-board sensors the CPU was hovering close to 61C and the thermometer was reading 41C from the heat sink.
It’s a close call, it appears that the toothpaste performed slightly better than the thermal compound. The data is so close that I am sure if I ran more trials the results from both products would converge to the point where there would be no noticeable difference between the two. So if you are really desperate I guess you could temporarily use toothpaste as an effective substitute for thermal compound. I would not recommend this but it seems that it does work.