A few weeks ago Arnie released a big thermalpaste roundup. If youâ€™ll recall, the OCZ Ultra paste turned in just average performance, not bad, but not particularly good either. In response to this, OCZ sent me their latest creation, OCZ Ultra II. While I donâ€™t have the 7 different compounds that Arnie had in the test, I do happen to have the excellent Arctic Silver III, as well as some original OCZ Ultra. Does the sequel have what it takes to take down AS III?
In case a few of your thermalpaste geeks out there are curious, I'll go into a little brief history on the popular kid on the block, Arctic Silver. The original AS I had around 4.65 to 5.15 W/m*K thermal conductivity depending on the version and particular batch. Doesnt exactly blow anyone's skirt up these days, but for that time it was an impressive compound. AS then introduced a greatly improved version II, made with 99.9% pure micronized silver and thermal conductivity of around 8.0 W/mK. AS III was a less dramatic improvement. Thermal conductivity increased to >9.0 W/mK.
OCZ's Ultra is the underdog in this test. It doesn't have the reputation of everyone's favorite Arctic Silver, but that's certainly not something to hold against it. The original paste wasn't spectacular, but OCZ has gone back to the drawing board and updated it. The new Ultra II, like AS II and III, is made with 99.9% pure micronized silver, and has a thermal conductivity of 8.0W/mK, which at least on paper puts it square into AS II territory. As you'll see in the benchmarks on the following page however, paper specs aren't the entire story here.
In order to be as consistent and accurate as possible, I followed Arnie's lead and used one of Penn Tech's temperature controlled net labs for NT students such as yours truly, so the ambient temp remained a constant 22 degrees celsius. I also used his method for cleaning the processor off before using a different compound. Arctic Silver states that 72 hours are required for their paste to fully "break-in", so I recorded results for 3 days. Remember however, that Arnie's specs differ from mine, so don't compare the results of our tests directly.