EK - Asus TUF 3080Ti Block

[LLNEO13]

Hello,

I was wondering if anyone has installed the EK waterblock for ASUS Tuf 3080Ti.

Although at full load the card reaches 60C it crashes after gaming a bit.

Switching the profile to quiet (and as a result effectively downcloaking the card a bit) solves the problem but as you can imagine it still leaves me troubled.

Both the waterblock, backplate and radiator are hot in touch and since the card was working prior to introducing it into the system leads me to excluse the psu (Corsair 860w) as a potential issue

I have 3 main candidates as the source of the failure
a) GPU and CPU are on seperate loops. When I was filling the gpu it did look like the flow was slow. I have used a couple of ek offset fittings. Dont know if they can restrain the flow that much

b) maybe there is poor contact? But if that is the case then the block and backplate wouldnt be hot when gaming would they + the temp would go well above 60C

c) poor contact in some of the chips on the back?

EK has not bothered to amend their instructions after the release of 3080ti and the online manual only shows the scenarios for 3080 and 3090.

https://www.ekwb.com/shop/EK-IM/EK-IM-3831109832639.pdf

I went by the instructions for 3080 but ultimately I dont think that it makes any difference as in both scenarios the pads on the areas that change are of the same thickness.

In any case I am going on holidays at the end of this week and I planning to adress all these scenarios by using the ek distro block and the ek active back plate. But in any case I wanted to check if someone else has encountered a similar issue

 

[MikeTimbers]

While 60C for a water-cooled GPU isn't stellar, it's not terrible. What radiator have you got on the GPU loop? What fans? Are the fans at full power? If not, can you put them on maximum and see if the cooling is better?

If the block is getting "hot" then the loop isn't performing overly well. The heat should be getting into the water and blown off the surface of the radiator.

 

[Ross Thomson]

I'd check to see if the block is making good contact with the die.

 

[stooeh]

If rad is getting hot then presumably the heat is making its way there okay. Run GPU-Z and make sure the Hotspot/VRAM temps are okay?

When I originally setup my loop I noticed my rad was getting a bit hot and it was mainly due to me not setting fan curves properly and they were capped at 40%

 

[Jay85]

Depending what radiator you're using, it might not be enough. I found that a single standard size 30mm 360 radiator wasn't great. Temps were around 55-60c and yes it was hot to the touch like you said.

I ended up with 2 x 55mm 360 radiators and that brought the temps right down. I'm now sat around 44c and the radiators are not as hot as the heat is being better dissipated with the extra rad space/fins etc.

Although you're problems should not really be temp related as 60c is still well within and below what a aircooled card would be. If its crashing, i would look up hwinfo and see if any other temps are getting higher than they should be. Check your memory junction temps and hot spot temps.

 

[LLNEO13]

Hello,

Thanks everyone for contributing

So the below temps are with
D5 at max rpm
EK PE 360 radiator
3x LianLi UL Unifans (the delta between full rpm and 75% is just 1.5C)

Max Values after 30 minutes of gaming (it happened that nothing crashed this time around, but still I find the temps high, I do believe that I had better temps with the tuf heatsink on)
GPU Clock [MHz] 1965
Memory Clock [MHz] 1187.7
GPU Temperature [°C] 59
Hot Spot [°C] 67.4
Memory Temperature [°C] 66

I am almost sure that something is going on with the pump, since as I have already mentioned the flow seemed low when I was filling up the loop. Just for my own peace of mind when I redo my loop ( I am going on holidays + to get married at the end of this week) in a months time I will reapply fresh thermal pads in the mem chips and all.