ASUS ROG Rampage III Extreme LGA1366 Motherboard Preview

Liquid Nitrogen Results

Next up is a look at how the board handles LN2. These results were from our first session with this particular chip under LN2 so it was a bit of an exploratory mission. First though, we needed to make a small change to the setup.

Obviously we needed to strap our KingpinCooling.com F1EE pot and swap in appropriate insulation. The difference between our phase and LN2 insulation is simply the top layer of Armacell. Once we have the new layer in place, we can mount the pot. We begin the session with some 2D results.

Super Pi 1M click for full size...	PiFast click for full size...
Super Pi 32M click for full size...	wPrime click for full size...

Well the reality of extreme benching hit us hard right out of the gates...cold bug. This chip we have is particularly horrible in this regard and really is the limiting factor in the benching we are going to see today. Being limited to a -85C cold bug really restricts what one can do with a chip, in our case it didn't instill much confidence to go over 1.80v for vCORE. Now that seems high already, but B1 retail stepping Gulftown CPU's really seem to like voltage: up to 2v in some instances. Our results are quite solid for being limited to 5.7GHz or less for clocks.

At this point, the board has worked flawless. Recovery from cold bug was great, just a matter of getting the pot to the right temp and rebooting. Overall, the board felt like we have been benching at these temps for a long time. The way the board reacted was 100% what we were looking for.

Unfortunately the cold bug is just part of life, this is where the CPU was most happy, -130C would have better, but that is just how it works some times. We have subsequently moved this processor into the EVGA X58 Classified and found it to initiate cold bug even sooner, around the -65C mark. This definitely means it is the CPU as the culprit, and also points to the R3E actually being quite good for holding off cold bug. A colleague of ours has found the exact same thing, lower cold bug on the R3E than the Classified. It is still early to draw conclusions but the data is starting to come in and it looks good fro the R3E.

AM3 click for full size...	3DMark 03 click for full size...
3DMark 05 click for full size...	3DMark 06 click for full size...

The 3D numbers really are strong for the clocks we are at. I personally wanted to clip 45K in 05 but with the CPU clocks we were limited to it just wasn't going to let that happen. We have found that this board really likes the 24X uncore multi. Which drastically improved uncore clocking over the 20X multi we were working with on phase cooling. Other than that, the board all but let us run what we wanted. Like the 2D results, cold bug was the single limiting factor and the board really let us max out our results given this limitation.