Nice video really makes me want to water cool my xtx Do you got any videos of real game performance and temps cause some games get my 7900 xtx hotter than timespy does
I know this is a older video but I feel like you really missed the opportunity to go with some K5 Pro here. Repasting a GPU should always swap to K5 if you're waterblocking it. Might as well lower the temps more right? Would be hard to spend like 25 extra bucks for the best thermal pad paste.
Couldve done alot of things better, since I basically added the water-cooling to an already existing system. Next one's going to be watercooled from the start and hopefully alot better looking. IT does work really well though, that's why I don't want to change anything-
Good job dude, nice temps. Can you share the hotspot temps and your stable gpu clocks ingame? I have the xfx merc 310 7900 xtx and on air I get 2850 mhz clocks in WZ2. Hoping to push it to 3000 mhz with my WB from Alphacool
I just played two rounds of Darktide, and got the GPU Clock on 2893 max, and junction temp: 70. The logging didn't work, so those are the stats that I saw while looking at the overlay. GPU Clock max might have been higher.
Good choice, the cooler on the Red Devil sucks. So damn heavy and big and for what. A card almost falling apart from its own weight. Really regretting not getting a reference card instead. Thinking of going water too
I have the non radeon edition and my hotspot temps still get insanely high with 2x 360mm radiators... undecided if i wanna tear it down and try a remount
What torque value did you used on that electrical screw driver? 0.6nm? higher or lower? It should say in the manual...If so what torque setting did you used to tighten the screws on your GPU PCB?
Here, I looked it up for you. You'll have to check the details yourself, but this is the one I'm using: www.startech.com/en-ca/computer-parts/ctk55pcedrive
Thank you very much for the reply i looked up the specification on the torque values of your electrical screw driver excerpt: "Battery operated forward and reverse modes with 0.25 to 0.35 N.m of torque" so here we have the torque values pretty weak i would do a bit more torque on the GPU waterblock for proper mounting pressure in the range of 0.4nm to max 0.6nm they might go higher but i wouldn't do it people used the EKWB torque screwdriver which is locked at 0.6nm and used it on their RTX 4090 waterblock without issue. However if you have GOOD gpu temps on GPU core, Memory junction and HOTSPOT temps then there is ZERO need to tighten the block why should you go through the hassle lol. Me...i think i tightened to LOOSELY and thus higher hotspot temps that is more so then on average water cooled rtx 4090 hotspot temps currently dissasemble my rig and re mount everything and its a PITA to do. Anyway thx for the reply. @@DanLinden
That's interesting! Fortunately I have very good temperatures and don't think its necessary. I might have given it a little twist myself in the end, with the screwdriver off. Can't remember.@@lawshadow2011
13:54 don't daisy chain the pcie connectors... each connector from the PSU is rated for 150W so you are splitting the load into 75w x2 not 150w x2 and the gpu gets 75w from the motherboard. the 7900xtx is rated for 355w so you gonna kill the gpu or the psu! good luck.
I double checked and you were actually right, it was one of those Y cables, with two at the end, but one in the PSU. Replaced it with two separate cables. Good observation!
this is totally wrong information. a dual 8 pin PCI-e cable can provide up to 225w + 150w on another separate cable + 75w from the PCI-e slot is 450W. These cards never go above 375w-380w at 100% load. You have at least 70W left in a daisy chain config. Many ATX 3.0 PSUs don't even offer three 8 pin PCI-e connections to be able to run three separate PCI-e cables. Do you have more headroom with 3 separate cables sure (525W), does that mean using 1 daisy chain cable is bad, no.
Dude wear some gloves. I don't think you realize how acidic fingerprints are...they definitely corrode metals on the PCB, etch them, and can even etch glass if left for long enough. Also give a surface for dust to stick to.
![[LIVE] : ONE ลุมพินี 89 | คู่เอก "ยอดไอคิว vs คิริลล์"](http://i.ytimg.com/vi/Z3vYkHJ_5Ig/mqdefault.jpg)
Brave using drill tightening.
The screwdriver is made for electronics and has a pretty weak torque. Don't think it can damage the board.
Nice video really makes me want to water cool my xtx
Do you got any videos of real game performance and temps cause some games get my 7900 xtx hotter than timespy does
I got the standard version, I think I like yours more since it is shorter and slightly thinner
I know this is a older video but I feel like you really missed the opportunity to go with some K5 Pro here. Repasting a GPU should always swap to K5 if you're waterblocking it. Might as well lower the temps more right? Would be hard to spend like 25 extra bucks for the best thermal pad paste.
Couldve done alot of things better, since I basically added the water-cooling to an already existing system. Next one's going to be watercooled from the start and hopefully alot better looking. IT does work really well though, that's why I don't want to change anything-
Like a parent changing a newborn's diaper.
Good job dude, nice temps. Can you share the hotspot temps and your stable gpu clocks ingame? I have the xfx merc 310 7900 xtx and on air I get 2850 mhz clocks in WZ2. Hoping to push it to 3000 mhz with my WB from Alphacool
I just played two rounds of Darktide, and got the GPU Clock on 2893 max, and junction temp: 70. The logging didn't work, so those are the stats that I saw while looking at the overlay. GPU Clock max might have been higher.
I'm putting my Red Devil 7900XTX in a block this weekend. It's the last piece to go into the loop. My rig will be fully wet.
Good choice, the cooler on the Red Devil sucks. So damn heavy and big and for what. A card almost falling apart from its own weight. Really regretting not getting a reference card instead. Thinking of going water too
I have the non radeon edition and my hotspot temps still get insanely high with 2x 360mm radiators... undecided if i wanna tear it down and try a remount
The only way to rule out that its your fault is to tear it down. Or you could try an internet search. Maybe its normal for that specific edition.
What torque value did you used on that electrical screw driver? 0.6nm? higher or lower? It should say in the manual...If so what torque setting did you used to tighten the screws on your GPU PCB?
Here, I looked it up for you. You'll have to check the details yourself, but this is the one I'm using: www.startech.com/en-ca/computer-parts/ctk55pcedrive
Thank you very much for the reply i looked up the specification on the torque values of your electrical screw driver excerpt: "Battery operated forward and reverse modes with 0.25 to 0.35 N.m of torque" so here we have the torque values pretty weak i would do a bit more torque on the GPU waterblock for proper mounting pressure in the range of 0.4nm to max 0.6nm they might go higher but i wouldn't do it people used the EKWB torque screwdriver which is locked at 0.6nm and used it on their RTX 4090 waterblock without issue. However if you have GOOD gpu temps on GPU core, Memory junction and HOTSPOT temps then there is ZERO need to tighten the block why should you go through the hassle lol. Me...i think i tightened to LOOSELY and thus higher hotspot temps that is more so then on average water cooled rtx 4090 hotspot temps currently dissasemble my rig and re mount everything and its a PITA to do. Anyway thx for the reply. @@DanLinden
That's interesting! Fortunately I have very good temperatures and don't think its necessary. I might have given it a little twist myself in the end, with the screwdriver off. Can't remember.@@lawshadow2011
Did you choose to have the acrylic part facing down?
I dont think you can put it in any other way. I don't like it either.
13:54 don't daisy chain the pcie connectors... each connector from the PSU is rated for 150W so you are splitting the load into 75w x2 not 150w x2 and the gpu gets 75w from the motherboard. the 7900xtx is rated for 355w so you gonna kill the gpu or the psu! good luck.
Hey! They're not daisy chained. It's to cables that go into the GPU 😄
I double checked and you were actually right, it was one of those Y cables, with two at the end, but one in the PSU. Replaced it with two separate cables. Good observation!
@@DanLinden you're welcome 😊
@@DanLinden It's a 400 watt card, so it's good you caught it.
this is totally wrong information. a dual 8 pin PCI-e cable can provide up to 225w + 150w on another separate cable + 75w from the PCI-e slot is 450W. These cards never go above 375w-380w at 100% load. You have at least 70W left in a daisy chain config. Many ATX 3.0 PSUs don't even offer three 8 pin PCI-e connections to be able to run three separate PCI-e cables. Do you have more headroom with 3 separate cables sure (525W), does that mean using 1 daisy chain cable is bad, no.
Dude wear some gloves. I don't think you realize how acidic fingerprints are...they definitely corrode metals on the PCB, etch them, and can even etch glass if left for long enough. Also give a surface for dust to stick to.