Heat is a major issue in today computing. With core speeds going higher and higher it's clear that we have to use powerful CPU coolers with better designed heatsinks and faster fans. If you read the web cooler reviews you have probably noticed the different results every site got with the same cooler. There are several reasons for that and the most important are:

• motherboard sensor. Most people read the cooler temperature from the motherboard sensor. It's not accurate, but it's a good choice in comparisons. Even two motherboards from the same model may not read the same temperature! Personally I agree with this method of testing but results must be considered only in the same article. A comparison with another cooler on another motherboard is irrelevant.
• different CPUs / Voltages. As you know the higher the frequency and voltage, more heat will result.
• different ambient temperature / humidity. Ambient temperature is important. A cooler temperature can be read when the case is opened or when the case is closed, so that you can not accurate compare your mobo temperature reading with a friend who has another case.
• different interface material. There are a lot of interface materials on the market: thermal grease, silicon thermal compound, silicon free thermal compound, etc. There are also a lot of manufacturers and the electrical conductivity is different.
• differences in the interface material installation. That's the point I wanted to reach. Let's see why.

The difference

In my last coolers roundup I pointed to a problem very popular between coolers on the market: the lack of installation instructions. Most readers will tell me that's not a big deal because there are not many users so ignorant not to know how to install a cooler. The truth is that it's not a big deal to install a cooler, but there are small things which can improve the heat transfer, ignored or not known by many users.

In our laboratories at PC Hardware we discovered major differences between a correct and an improper interface material installation. We conducted several tests with a Duron 700Mhz to see the importance of the interface material. The results proved us that without any interface material the CPU can easily burn even with a high quality cooler! So, without a thermal interface the CPU reached 62C, with improper installed thermal grease it reached 48C and with proper installed thermal grease the temperature reading was around 44C. Take notice that we used a Duron, which does not have the power requirements of a Tbird and we also used thermal grease, not thermal compound. I didn't use a Thunderbird mainly because I didn't want to risk a CPU crash in the no thermal interface test.

In the following we will describe a correct interface material installation and I strongly suggest you to use it because you can reduce your CPU temperature with several degrees at no money investment! With a correct thermal interface installation you can achieve the same difference like the difference between the thermal grease and the thermal compound. For more information about coolers you can read this article.

1. I suppose that you already have thermal grease or compound. If you bought a cooler with Chomerics T725 Phase Change thermal pad installed be aware that you can use the T725 interface only once, when you change your CPU you must clean the previous thermal interface and apply another thermal interface. That's why I suggest the use of Phase Change material on OEM market and I recommend users to purchase a syringe with thermal compound which can be used several times. In the picture in the left you can see how a CPU with unclean Phase Change material looks. Do not use the CPU in this state! Clean it first.