

Idle temperatures were taken ten minutes after each cold boot max-load runs were all twenty minutes long (I observed that temperatures didn’t tend to increase at all past the ten minute mark using MSI Kombustor). I did each test five times, with a cool-down period of ten minutes each between runs augmented with an electric fan cooling down the test machine after each test. It’s hard to say whether or not I could get improved temperatures if I ran this GTX 680 on its own water loop, but the fact that the one fitted with the NZXT Kraken is on a par with a custom water cooling solution is hugely impressive.Įach temperature test run was conducted at the reported ambient temperature (72☏/22.22☌) each time, controlled through my apartment’s central temperature controls. Run through similar conditions (same ambient temperature, same testing procedures), the best temperature I got for the GPU temperature is 96.8☏/36☌.
#Msi kombustor settings gtx 770 full#
The GPU has an EK EK-FC680 Acetal + Nickel full coverage block. This GTX 680 is part of my custom water loop, where both the GPU and the CPU are cooled with an XSPC RX360 radiator fitted with six Scythe Gentle Typhoons in a push-pull configuration. Just to put these numbers into some kind of context, I have another GTX 680 in my everyday PC. With the NZXT X40 + G10 combo installed, however, the highest temperature the GTX 680 hit was a mere 95☏/35☌! That’s an amazing 114.8☏/46☌ improvement over stock cooling. The delta temperature between ambient and GPU was 105.8☏/58.78☌. With ambient temperature kept constant (72☏/22.22☌), the GTX 680’s maximum load temperature hit 177.8☏/81☌. These screenshots were taken during a MSI Kombustor GPU stress test run (a 20min run each time). GeForce GTX 680 w/ Kraken + X40 Cooling (Load)

The smaller the difference between ambient temperatures and chip core temperatures (in this case, the GPU’s), the more efficient the cooling solution is.Īnd things are, if anything, even more impressive when you put the GPU through a maximum load condition. The 11☌ improvement by itself at idle conditions is hugely impressive, but I’m far more impressed with the temperature delta between ambient temperatures and the GPU’s temperature at idle (a difference of 5☏/2.78☌). In comparison, at an identical ambient temperature and given the same conditions (10 minute idle period post-cold boot), the GTX 680 cooled by NZXT’s X40 registered 77☏/25☌. These temperatures were taken with the test machine left on idle (no user input) for ten minutes after a cold boot. GeForce GTX 680 w/ Kraken + X40 Cooling (Idle)Īt an ambient temperature of 72☏/22.22☌, the GTX 680 with its stock cooler stabilized at 96.8☏/36☌. I think these testing results speak for themselves. The whole point, though, of using the Kraken G10 is to see whether or not adapting an Asetek-designed CPU AIO water cooler for use on a GPU is a performance enhancement. I’d say that if I had to do things over, I could perform the installation in half the time.īut ease of installation is just one part of the equation. This time includes removal of the stock cooler. All told, going carefully and checking fits a few times before fastening things down, installation of the NZXT X40 onto the GTX 680 using the G10 bracket took about an hour and a half. The diagrams are well-drawn and are augmented with basic textual instructions. NZXT’s installation instructions for its G10 are easy to follow.
