The Nanotherm product comes in two different varieties; Nanotherm Ice and Nanotherm Blue. The "Ice" variety is its natural form and is an off-white color. The "Blue" variety has had a ceramic coloring agent added to make it a light shade of blue. Both types were tested and the results were the same, so I'll just report results for the Nanotherm line in general.
For those who have used Arctic Silver, you know that it is pretty thick when applied to the core of the processor. This compound had a lighter consistency and didn't seem quite as heavy as the AS product. It applied very easily, and the texture was very smooth. It also cleaned up much easier than the AS goop. This is a huge plus when you change over heatsinks a lot like I do.
So what are we waiting for? Let's see how it held up…
Testing was conducted on my standard system that is used for heatsinks. The sink used for this test is the OCZ Gladiator. Since we already know how well it performs, we can skip the low-end speeds and focus on an aggressive overclock. This should allow us to see how well the new thermal compound holds up on "Power User" system. The test system consists of:
Antec SX1030 Tower Case
AMD Thunderbird 1000 (AVIA)
Abit KT7A-RAID Motherboard
256MB Crucial PC133 CAS2 SDRAM
Arctic Silver Thermal Compound
Creative Annihilator 2 GeForce2 GTS
The Antec case houses four 80mm fans for a total of 170-CFM airflow moving through the case. The Gladiator HSF has the optional Delta fan mounted, and the processor is set to 1333MHz. I'm going to try to cook this thermal compound right off the face of the processor so that we can see if it will be acceptable for use in an overclocked rig. There's no sense in half-stepping now and then cooking a chip later on.
Testing itself was also changed up a bit from my heatsink tests. Since we know the capabilities of the sink itself, I wanted to see what happened with the interface material at idle, after a Deathmatch of Quake III Arena, and at 100% processor usage. The final test was conducted by running the UD Cancer Research program for 20 uninterrupted minutes. Since this program uses all idle CPU clock cycles, its end result is a 100% usage when there is no other activity going on.
So how did the new goop perform?
Not too shabby at all. The Nanotherm struggled just a bit with the 100% CPU usage, but it's results are still very good considering a 33% overclock on the processor and full clock cycles being gobbled up by the UD program. And considering that the cost is estimated at being considerably less than the Arctic Silver line, it is really making this a choice to consider.