Temps have gotten a little toasty inside my H440 with two GTX 980 Ti and a 750 Ti, so I've begun work on dropping temps. Step one, was to swap out the stock fans on the case and Kraken X61 with Noctuas NF series industrial PPC 3000 RPM fans. While this dropped my CPU load temps by a fair amount (81c down to 69-73c at 100% load in stress tests; I also managed to slightly lower voltages from the auto overclock settings), I am still left with a pair of cards running at 79c under load (with priority given to temps at 79c vs power in afterburner) and 40-45c at idle which causes my CPU to also idle right around 45c with the Kraken X61 pulling in that hot exhaust.
Step two is to try and fix the GPU temps and their associated fan noise under load. Let me tell you, the fan at 70% is loud! Don't even get me started when it's going full speed; though it does drop temps from 79 down to around 72c and 75c on cards 1 and 2. This has brought me to the conclusion that the ACX cooler just doesn't cut it when run in SLI, and as such, I'm taking my ACX coolers off my cards (980 Ti FTW) and installing the blocks from the hybrid kit, but won't have a fan from the reference cooler directing air towards the VRMs due to compatibility issues with my cards semi-custom design. I've looked at modding in the reference cooler, but there appears to be a number of issues, so I've moved towards just installing the water block.
My assumption, is that without the 80c heat from the aluminum ACX heatsink hovering over the cooling-plate (along with the downdraft of hot-air exhuast); coupled with cooler air coming off the front mounted Noctua NF-F12s, there should be enough airflow to keep the the VRM temps at least in check. Does anyone know what their limit is before permanent damage occurs? I've looked at aluminum/copper heatsinks, as well as thermalpads to place over the cooling plate, but hear that they have a tendency to fall off due to the adhesive bond weakening under high temps coupled with gravitational pull. Any suggestions, would be greatly appreciated!
Step two is to try and fix the GPU temps and their associated fan noise under load. Let me tell you, the fan at 70% is loud! Don't even get me started when it's going full speed; though it does drop temps from 79 down to around 72c and 75c on cards 1 and 2. This has brought me to the conclusion that the ACX cooler just doesn't cut it when run in SLI, and as such, I'm taking my ACX coolers off my cards (980 Ti FTW) and installing the blocks from the hybrid kit, but won't have a fan from the reference cooler directing air towards the VRMs due to compatibility issues with my cards semi-custom design. I've looked at modding in the reference cooler, but there appears to be a number of issues, so I've moved towards just installing the water block.
My assumption, is that without the 80c heat from the aluminum ACX heatsink hovering over the cooling-plate (along with the downdraft of hot-air exhuast); coupled with cooler air coming off the front mounted Noctua NF-F12s, there should be enough airflow to keep the the VRM temps at least in check. Does anyone know what their limit is before permanent damage occurs? I've looked at aluminum/copper heatsinks, as well as thermalpads to place over the cooling plate, but hear that they have a tendency to fall off due to the adhesive bond weakening under high temps coupled with gravitational pull. Any suggestions, would be greatly appreciated!