I've had my cooler installed on this 3090 for a week now. I have did the firmware update and tried both the stock and oc firmware (not the beta). X1 see's it as a Hyrbid. I have the cooler mounted at the top as the instructions have suggested for best performance. I have tried some Thermaltake Riing Trios but those fans actually don't see to perform as well even at full speed so I'm back to the stock fans that came with the evga AIO cooler. 
 
I'm getting temps all over the place and I know I installed everything correctly, took my time, etc. I used the paste that was on the block when I got it, I really do not think changing that is going to solve my issues here as these temps don't seem ideal for a cooler. If I run Time Spy test the temps are not horrible but in real world gaming, like today in Path of Exile I'm getting temps hitting around 70c and on GPU2 even higher. This is on the stock firmware clock and the switch off for OC on the card itself. With OCing it was even worse.
 
Please let me know what if any other details you need. I would love to figure this out here before reaching out to EVGA support. If you have this same card and AIO from evga I would love to hear your temps in games. I will say that my temps vary from game to game. I just did not expect to see temps getting over 60c with anything thrown at this card with this AIO installed. Am I expecting too much?
 
 

Do you have any suggestions?

You installed the cooler yourself, every time someone has had temps like this after a hybrid install it's been the mount that's the issue, they've either used the wrong size thermal pads near the vRAM, or something else is just not quite right.  It doesn't take much to see higher than normal temps.
Unless your ambient is insanely high there's no way what so ever your load temps should be in the 70s.  At the most you should be occasionally hitting 60 if that.
For comparison, my load & idle temps:
 

Note that's with a pretty decent undervolt.  I've started running 2040MHz @ 1000mV instead & it runs about the same temps.
 

And that's idles.  If your idles aren't at least close to that, your mount is off... this is with ambient in the 70F-72F range.

GPU2 is on the back of the card, in the cutout of the backplate. Put a fan over it if you want it cooler.

As to GPU1 hitting 70C, that is high. Hopefully a repaste and a reseat will fix it. Mine gets to 64C pulling about 500W (stock is 450W), but I've got really good case airflow and high static pressure corsair ML120 fans.

Edit: don't be afraid to set an aggressive fan profile.

Ok, I'm back and while I do have some better results they are not satisfactory. I ran Time Spy and the HWInfo info is here, max 63c (better but not perfect). I did notice something and I will take blame for it but the instructions are not clear on this and I have attached pictures. There is some thermal pads to the sides of these areas marked in red and I assumed I was not suppose to take them off as they are not marked in red. This is the pic of the card before I started install and you can see thermal pads there. Well turns out the AIO had some so they got double stacked.  Would that be enough to get my temps to go up on gpu? All the other thermal pads are correct. I did apply new thermal paste to the gpu (MX-4).
 
If you you think those thermal pads on the sides (is that the memory it's over?) are the issue is there any recommended thermal pads I can purchase? 
 
Thanks for all the help. When I went back into Path of Exile and played for awhile I was getting temps in the low 60's 58-63 so it did improve some. This is how my setup is, I can remove the top (see the slip between lid) for more air flow but I don't think that should be required to get the ideal temps?

Thanks for the recommendation! I was going to come back here and post my logs and info you asked for but the temps looked great! Well sort of, I removed the top lid to get to those temps. Think I need a better case. Not sure if you saw in my other post the picture where it shows the lid of my case. Well that can be removed as it's magnetic. So for this I went ahead and ran it like this. Played my game for awhile (for specs I'm playing 1440p on a Samsung 32" 240Hz monitor with Gysnc on). Getting around 200FPS with maxed graphics. Not one time did it ever go over 54c. Now that is with 80% fan speed which imo seems a bit much as it's pretty loud. Maybe it would have been just as good at 60-70% but didn't try yet. After about 15 minutes of playing and not seeing it get over 54c I put the top back on but I left the gap like you see here. As soon as I put it on the temps went up to 64c, like with in less than 30 seconds it jumped 10c. If this top piece was mesh I would be good to go I think. So yeah, time to go look for a better case. Too bad I really liked this case as it was very modular and easy to work with.
 
I still wonder if these temps people are getting where they don't go over 54c are able to do that on stock an profiles, etc. Now that this seems to work better I'm going to keep the lid off and try some other games that don't push it as hard. Believe it or not Cyberpunk has better temps then Path of Exile. 

Rbk_3

Dabadger84

And that's idles.  If your idles aren't at least close to that, your mount is off... this is with ambient in the 70F-72F range.

Or if your idles are in the 30's, you may have Power management set to Prefer Maximum Performance. 

I do have it set to that! I'll see what happens when I move it back to default in nvidia panel.

Chad711

kevinc313
 
Sounds good, looks like you've got some pretty serious restriction with that top plate.
 
Keep an eye on your ram and vrm temps (MOS, or MOSFET) temps since you're running at high FPS, there may be an issue with then getting too much heat with the hybrid rad blowing on them, they may run 10C or so degrees hotter than they would otherwise.  If that's OK or not depends on how much temp headroom they have in your setup.

 
Where can I see this "MOS & MOSFET" you are mentioning? I do have HWInfo up if that is where I should be looking but I do not see these labels. I also have X1.
 
I have let my PC idle for 2+ hours and it seems these temps are a bit high for idle. This is with my lid off too. I did order a new case today, the Meshify 2. 

In HWiNFO on my Z390 machine MOS is about half way down under the motherboard heading, with system, m.2, pcie temps etc and the MB driven fan speeds.  MOS can get up over 100C, so keep an eye on it depending on what kind of load the CPU is under when gaming.  This is the VRM's driving the CPU. 
 
On the ICX temps, make sure your VRM fans is running fast enough on the card.  I've been using the OC switch setting bios, so the min speed is 40% or 1200 RPM. Everything is in the 20's at idle.

If you're not downclocking to 210mhz at idle, it probably won't idle in the 20's.

Chad711
Is the firmware suppose to downclock automatically? I don't overclock very much so just want to make sure I'm understanding.

Yeah if you have a single monitor and the Nvidia power management setting is on "normal" or "adaptive", it will downclock. "Prefer Maximum Performance" will stay at like 1800mhz, sometimes multiple monitor setups will cause it to not downclock.

Dabadger84

Chad711
Thanks I'll keep an eye on it. I think I might have a bad cooler here. My temps are never in the 20's at idle. The only time I saw 29c was today when I put the fans are 80% during idle and that was only on gpu1. Everything else higher.

Do you have multiple monitors or just the one that you stated previously?  I have my radiator fans go all the way down to 20% fan speed at idle, with the on board GPU fan turning off completely, & I'm seeing 22-24C idle depending on ambient room temperature.  
Right now my ambient is 72F (warm for this room) & the card is at 24C core, 29-30C on the vRAM, 30-32C on the VRM/PWR readouts.
Your card should be idling at 210MHz core, 101MHz vRAM, if it's not "downclocking" to that on Desktop, and you already turned off Prefer Maximum Performance, something else is going on.
What drivers are you running currently?  On the 460.89 drivers, I noticed that at my native resolution on this new monitor (5120x1440) at 240Hz, the card wouldn't always downclock fully on desktop, but at 120Hz it would.  I updated to the 461.09 drivers last night, and the card now idles fully downclocked even at the full 240Hz.  So it could also be a simple driver/idle issue.
 
Also if you could post a picture of your case's internals, that would help as well. 

I have two monitors, both 32" Samsung monitors, one is the older 144Hz and then main is 240Hz. I'm ono the current 461.09 drivers too. I have a new case coming this week (Meshify 2). This is a pic I posted earlier but I have since moved my fans to intake up top as Kevin suggested which helped more than anything so far.
 
I contacted support, they suggested I try something which I'm about to do here shortly. 
 
Their recommendation:
I always recommend an option where, as it looks like you are using a traditional style of case and mounting set up, to have the computer case turned on its side with the motherboard now laying horizontal and the graphics card now vertical. Let the system run like this for a few minutes and you may hear the pump gurgle or bubble during this time. This would confirm that there is more than likely air trapped in the pump area that needs to be cleared. If this is the case, you can shut down the system and next unmount the radiator from the chassis and elevate it physically above the GPU with the tubing going straight up to the radiator from the card, and the tubing going straight down the card. Let the system run again for a few minutes this way and then feel free to stand the case back up (being mindful to keep the radiator above the card at all times) and remount the radiator.