pclausen

Nereus
 
Will be interesting to see what difference it makes once the thermal tape is on. Have you looked under the back plate to see what EVGA put under there?

 
I have not removed the back plate, but I looked at a pic shows exactly where the VRAM is located.  So I changed my arrangement slightly after applying the thermal tape to be as follows:

 
I also upgraded to a Nuctua 3,000 rpm capable fan and slightly changed its location to direct a bit more air over the heat sinks closest to the motherboard.  I basically raised it slightly so that it was at a slight angle and rested on the CPU tightening knobs:
 

 
Turns out the thermal tape did not make much difference since the heat sinks made excellent contract and transferred heat quite well just resting on the back plate.
 
Here are my results with the fan at 2,000 (Current) and 2,900 rpm (Minimum):
 

 
I'm happy with the readings I get with the fan @ 2,000 rpm as it is relatively quiet where the 2,900 setting is annoying and only lowers my temp by 2 degrees max.

I did something similar with heatsinks direct contact with the backplate. Noctua fan on top or heatsinks. Reduced around 4-5°C.

So adding a high-end conductivity thermal pad between the heatsink and the backpkate won't help much reducing the temps? (I'll save around 40 bucks)

Temps will vary depending on the case and airflow setup.  If you're getting high temps and depending on the case, you'll either want direct airflow on the backplate in a open case or test a closed case with the door open and closed and reconfigure the fans for proper efficiency via positive pressure.  That also doesn't exclude ambient temps as you'll want the coolest air entering the case.

menko2
I did something similar with heatsinks direct contact with the backplate. Noctua fan on top or heatsinks. Reduced around 4-5°C.

So adding a high-end conductivity thermal pad between the heatsink and the backpkate won't help much reducing the temps? (I'll save around 40 bucks)

Currently I'm seeing 3-11C lower temps on my vRAM iCX sensors, and that's pre-redoing/improving thermal pads, just with the heatsinks only & a fan blowing over them:
 

And those temperature reductions are despite a higher Core OC/voltage & a +1000MHz memory OC.
 
If it weren't for me messing up my back on Sunday, I would have already done the thermal pads, but I can't lean/bend over things for any amount of time right now so operating on the backside of the GPU is out of the question for at least a few more days.

Dabadger84

menko2
I did something similar with heatsinks direct contact with the backplate. Noctua fan on top of heatsinks blowing air up (not sure if has to blowing up or down). Reduced around 4-5°C.

Used three heatsinks. Two 100x60(30mm tall) and one 100x100(20mm tall). My ram is low and I have to put a lower heatsink go cover that area below.

So adding a high-end conductivity thermal pad between the heatsink and the backpkate won't help much reducing the temps? (I'll save around 40 bucks)

Currently I'm seeing 3-11C lower temps on my vRAM iCX sensors, and that's pre-redoing/improving thermal pads, just with the heatsinks only & a fan blowing over them:
 

And those temperature reductions are despite a higher Core OC/voltage & a +1000MHz memory OC.
 
If it weren't for me messing up my back on Sunday, I would have already done the thermal pads, but I can't lean/bend over things for any amount of time right now so operating on the backside of the GPU is out of the question for at least a few more days.

That's looking good. Even when you will do more work when your back is good again.

I won't open the card or back plate from stock.

As you can see my heatsinks are different. Two are 100x60x30mm and one 100x100x20mm because the ram blocks 30mm tall ones. My fans are blowing up.

My question is that if I put a 15k/w 0.5mm thermal pad between the backplate and heatsink will help or not much. I read in a post that it didn't help much to another person but it was a 6k/w thermal pad.

Do you know if it will make any effect?

 
My KP build as it stands. I need some time to figure out how to change the voltage settings on this card. Having issues passed 2145mhz on the core.

Got the KP installed today and the max power draw I can make is about 440w.  PCI-e pin 3 seems to never draw more than about 80w, and the pcie slot draws no more than 60w.  Card will run through timespy at 2200mhz solid the entire way at about 1.05v - but the scores are lower than a run at 2000mhz.  Doesn't seem to power throttle anywhere - at least gpuz doesn't report it that way.  Not entirely sure why the scores are lower - but I'll be testing it more over the next bit to see if I can find a solution.

menko2
That's looking good. Even when you will do more work when your back is good again.

I won't open the card or back plate from stock.

As you can see my heatsinks are different. Two are 100x60x30mm and one 100x100x20mm because the ram blocks 30mm tall ones. My fans are blowing up.

My question is that if I put a 15k/w 0.5mm thermal pad between the backplate and heatsink will help or not much. I read in a post that it didn't help much to another person but it was a 6k/w thermal pad.

Do you know if it will make any effect?

Is there nothing on there currently or Thermal tape?  It's possible doing Thermal Pads if they're thin like you say would help more than hurt.  I'd say pretty likely it would help, vs nothing and vs thermal tape - because from what I understand most thermal tape doesn't have too great of conductivity.

Hi everyone, I received my 3090 Kingpin hybrid a couple of weeks ago and have been having a really weird issue. Sometimes after a random cold start the AIO pump decides to not start. I noticed this first time when I shut down my computer to work on something and then started it back up. As soon as my computer booted up in windows, the hybrid fan as well as the GPU temp controlled radiator fans went crazy followed by an awful "GPU OVERHEAT" message. I checked x1 precision and the GPU temp was on 90°C and rising. I was on the LN2 bios which apparently doesn't have safeguards so I panicked and turned off the PSU.

After some more panicking, sweating and uninstalling and reinstalling the video card it suddenly started working.

Fast forward to today and I decided on updating my BIOS. Same ordeal again. This is ridiculous. Any tips?

Edit: After trying all sorts of things I gave up, let the card sit for an hour. Plugged everything in and the pump started working again. The easiest way to tell if the pump is working is the hoses start vibrating. Still need to find a permanent solution because this is annoying.

pcgamerv2
Hi everyone, I received my 3090 Kingpin hybrid a couple of weeks ago and have been having a really weird issue. Sometimes after a random cold start the AIO pump decides to not start. I noticed this first time when I shut down my computer to work on something and then started it back up. As soon as my computer booted up in windows, the hybrid fan as well as the GPU temp controlled radiator fans went crazy followed by an awful "GPU OVERHEAT" message. I checked x1 precision and the GPU temp was on 90°C and rising. I was on the LN2 bios which apparently doesn't have safeguards so I panicked and turned off the PSU.

After some more panicking, sweating and uninstalling and reinstalling the video card it suddenly started working.

Fast forward to today and I decided on updating my BIOS. Same ordeal again. This is ridiculous. Any tips?

Edit: After trying all sorts of things I gave up, let the card sit for an hour. Plugged everything in and the pump started working again. The easiest way to tell if the pump is working is the hoses start vibrating. Still need to find a permanent solution because this is annoying.

The stock "LN2" 520W BIOS does still have protections in place - the card would've shut down automatically if it got any hotter.  You have to download the 1000W BIOS & Classified Tool to get a "fully unlocked" BIOS & that technically voids your warranty if they find out you used it.
 
I would suggest contacting eVGA Support/Customer Service (highly recommend doing so via phone) & seeing what they recommend, that pump issue should not be happening and it may indicate either the fan header for the pump is lose (rare, but it happens) or the pump is bad possibly.  They may ask if you're comfortable with taking the card apart enough to check if the plug in is loose or not - if you're not, RMAing it may be the only fix, if you are, I'm sure they can walk you through it.

Dabadger84

menko2
That's looking good. Even when you will do more work when your back is good again.

I won't open the card or back plate from stock.

As you can see my heatsinks are different. Two are 100x60x30mm and one 100x100x20mm because the ram blocks 30mm tall ones. My fans are blowing up.

My question is that if I put a 15k/w 0.5mm thermal pad between the backplate and heatsink will help or not much. I read in a post that it didn't help much to another person but it was a 6k/w thermal pad.

Do you know if it will make any effect?

Is there nothing on there currently or Thermal tape?  It's possible doing Thermal Pads if they're thin like you say would help more than hurt.  I'd say pretty likely it would help, vs nothing and vs thermal tape - because from what I understand most thermal tape doesn't have too great of conductivity.

Exactly. Right now there is nothing between the heatsink and the backplate.

The heatsinks get hot so there is some heat transmition for sure.

6k/w thermal pads are cheap but the 15k/w are expensive.

Will be very noticeable the different thermal pads in this particular case?