Hi,
Over 2018 I built up a PC, modified it on air and then slowly moved it over to water. I thought I'd share some details of what I built and some of the temperature data (I'll add that as I go because I don't have all to hand except GPU) as it may be interesting for others. I'm open to try running any test and share the data for any comparisons people would like.
Most relevant specs (if you want more, just ask!)
Case: Fractal Meshify C
System: 8700k. 5ghz @ 1.38v. De-lid and replacement copper "rocket" IHS.
Mobo: Z370M D3H (rev. 1.0) (it came with the package - VRMs not ideal but it works!)
RAM: 32mb corsair vengeance 3333mhz. (for some reason it seems stuck at 2840mhz or so regardless of XMP etc).
GPU: 1080ti Zotac mini. Overclocked to hold a constant non throttling 2062mhz and memory of 5600ghz (rounded). It can run more than 2062 but you see the odd bit of throttling during extended benchmarking so it is pulled back for stability for gaming.
CPU Cooling: Running the Alphacool XP3 waterblock that comes with the Alphacool Eiswand 360 kit.
GPU Cooling: Alphacool GPX waterblock (the one with built in pump) and with finned backplate and lots of fins to cover VRM cooling etc. Liquid metal used.
Radiators: Alphacool Eiswand 360mm external tower radiator (360mm Alphacool NexXxos xt45mm) with the standard fans replaced with 6x Corsair High Performance static pressure 120mm. In addition a 240mm Alphacool NexXxos xt45mm radiator with 4x Corsair ML Pro 120 on push/pull. This is in the front of the Meshify C.
Pumps: All in series - 2 x DC-LT 2400rpm that are built into the base of the Eiswand 360 tower and 1 x the suction pump that is built into the GPU block.
Loop: 11/8mm hosing that came with the Eiswand plus some more that was needed to add in the GPU. All quick release fastenings removed. Just straight through pipe. Exits the 360mm tower -> CPU block -> 240mm radiator -> GPU block -> back into 360mm tower.
Interesting point:
The data that I thought was interesting is that the GPU under 100% load for hours never runs more than 7 to 8 degrees above room temperature. This is with the fans on the 360mm and the 240mm all turned up and pumps working at 100%. This is during gaming or benchmarking. So the CPU is seeing some load just not 100%. If you load up the CPU to 100% it doesn't really seem to effect the GPU temps but I find it hard to get a real 100% load on GPU and CPU at the same time to test for a longer period.
I believe the DC-LT pumps are perceived as being weak (which I think they are compared to strong D5 etc?) but the performance is there.
Welcome any points etc. I can share some photos. The system isn't built for beauty but the performance is there imo.
Another bit of info. I can run the system with the tower turned off (fans and pumps) and just use the GPU pump and the 240mm fans. Fine for browsing etc. It actually is fine for playing less stressful games like NIOH but for PUBG (the regular game of choice) you want to run the tower aswell if you want very low temps. I could test it without the tower for whatever tests people want.
I'm in the middle of building up a Lian Li PC-011 and moving all cooling internal. Open to share data of that build too! Cheers all.
Steven_RW
Over 2018 I built up a PC, modified it on air and then slowly moved it over to water. I thought I'd share some details of what I built and some of the temperature data (I'll add that as I go because I don't have all to hand except GPU) as it may be interesting for others. I'm open to try running any test and share the data for any comparisons people would like.
Most relevant specs (if you want more, just ask!)
Case: Fractal Meshify C
System: 8700k. 5ghz @ 1.38v. De-lid and replacement copper "rocket" IHS.
Mobo: Z370M D3H (rev. 1.0) (it came with the package - VRMs not ideal but it works!)
RAM: 32mb corsair vengeance 3333mhz. (for some reason it seems stuck at 2840mhz or so regardless of XMP etc).
GPU: 1080ti Zotac mini. Overclocked to hold a constant non throttling 2062mhz and memory of 5600ghz (rounded). It can run more than 2062 but you see the odd bit of throttling during extended benchmarking so it is pulled back for stability for gaming.
CPU Cooling: Running the Alphacool XP3 waterblock that comes with the Alphacool Eiswand 360 kit.
GPU Cooling: Alphacool GPX waterblock (the one with built in pump) and with finned backplate and lots of fins to cover VRM cooling etc. Liquid metal used.
Radiators: Alphacool Eiswand 360mm external tower radiator (360mm Alphacool NexXxos xt45mm) with the standard fans replaced with 6x Corsair High Performance static pressure 120mm. In addition a 240mm Alphacool NexXxos xt45mm radiator with 4x Corsair ML Pro 120 on push/pull. This is in the front of the Meshify C.
Pumps: All in series - 2 x DC-LT 2400rpm that are built into the base of the Eiswand 360 tower and 1 x the suction pump that is built into the GPU block.
Loop: 11/8mm hosing that came with the Eiswand plus some more that was needed to add in the GPU. All quick release fastenings removed. Just straight through pipe. Exits the 360mm tower -> CPU block -> 240mm radiator -> GPU block -> back into 360mm tower.
Interesting point:
The data that I thought was interesting is that the GPU under 100% load for hours never runs more than 7 to 8 degrees above room temperature. This is with the fans on the 360mm and the 240mm all turned up and pumps working at 100%. This is during gaming or benchmarking. So the CPU is seeing some load just not 100%. If you load up the CPU to 100% it doesn't really seem to effect the GPU temps but I find it hard to get a real 100% load on GPU and CPU at the same time to test for a longer period.
I believe the DC-LT pumps are perceived as being weak (which I think they are compared to strong D5 etc?) but the performance is there.
Welcome any points etc. I can share some photos. The system isn't built for beauty but the performance is there imo.
Another bit of info. I can run the system with the tower turned off (fans and pumps) and just use the GPU pump and the 240mm fans. Fine for browsing etc. It actually is fine for playing less stressful games like NIOH but for PUBG (the regular game of choice) you want to run the tower aswell if you want very low temps. I could test it without the tower for whatever tests people want.
I'm in the middle of building up a Lian Li PC-011 and moving all cooling internal. Open to share data of that build too! Cheers all.
Steven_RW