mrshrir
I dont understand how you guys are able to SET the voltage ? i have an asrock mobo . 

In the mobo bios, you go into the fan control header location you've plugged the new pump cable into, set it to DC voltage mode (not PWM), then set a volt or percent curve based on CPU or system temp. 
 
I tried CPU temp but switched over to a 35C system temp point where it kicks from start/idle at 7.32v to 10.5v loaded.  Same with the hybrid radiator fans being plugged in to the MB, going from 1000rpm to 1800rpm at 35C system temp, though they are PWM control.

Yes, and also you cant control pump via graphics card, so both should be controlled via motherboard.
 
Or you can get another mini 4pin - 4pin fan adapter like this:

and power up 4pin fan via card header.
 
OR there is 4pin pwm fan header on plate itself, but it can be controlled only by x1 software.

Is there any downside to using the magnet mod other than it being completely ghetto? I'm about to replace my 1080ti Hybrid (silent all the time) with an FTW3 Hybrid so I'm preparing to have to deal with this sound issue. Obviously I'll probably do the cable mod but was just curious as to why the magnet mod is frowned upon? 

DeadlyMercury

RipGroove
Is there any downside to using the magnet mod other than it being completely ghetto? I'm about to replace my 1080ti Hybrid (silent all the time) with an FTW3 Hybrid so I'm preparing to have to deal with this sound issue. Obviously I'll probably do the cable mod but was just curious as to why the magnet mod is frowned upon? 

Because it do the same thing (slows down impeller), but you have no control over it. So you can't say is your pump still running or not, cant control it as temperature function.

Well that makes sense.

 
Okay so there are guys here with way more experience than me but I thought I'd share my problem and how I solved it very easily.  If it doesn't work for you, I'm sorry for taking up your time.
 
So when I recently purchased a Corsair Hydro H55 the very first thing I did before getting it on my evga 2080 RTX super was simply plug the pump straight into the PUMP/FAN3 header (Set to silent in Bios), just to hear the noise.  As I held it in my hand with the heat sink part facing up (The way it will attach to the bottom of the cards gpu) it made a nasty buzzing/gurgly noise,  I tapped it, shook it, followed all the tricks online and nothing would make it go away.
 
So I turned it sideways.  Instantly it went %100 quiet.  Turned it back the way it was and the loud noise returned.  So of course solution 1 was:
 
1) Vertically mount the Video card so the pump would be facing sideways
 
Then I decided to see just how loud the pump would sound on full speed setting in the Bios, with the heatsink facing up again.  It TOTALLY went quiet.  I lowered the speeds and they all made the noise.
 
So I settled for
 
2) Keep the  pump at max speed at all times.  Dead silent.
 
 
So maybe vertical mount the GPU?   Hope this can help in some way.
 
Edit: In case it wasn't clear, I needed to use a Kraken G12 to mount the H55 to the 2080  (Must use AMD brackets, Nvidia ones won't fit for 2080s)

Yeah, the hybrid cooler I put on my 2080TI FTW3 Ultra is pretty annoying to hear. I bought it to get rid of the noise from the stock cooler while gaming and well, it is a little quieter and definitely cooler, buuuuuuut that pump noise is very grating to me. I knew of the complaints before I picked it up, but still... I don't see how this was ok to produce honestly. My 980TI Hybrid was a much quieter card and had zero pump noise it's entire life.

Nazarus
Turn it to hydro.  Problem solved!

I have one of these FTW3 Hybrid cards, a SUPER model.
 
This isn't a solution, especially to the people who paid a premium for a card expecting an improvement with respect to dB and temps. Switching to hydro or air is an additional purchase, and an insult to the customer base that EVGA are supposed to be supporting. I'm extremely surprised that they are not offering a pump/cooling replacement programme to take care of the issue, to at least show that they care about customer satisfaction/restoring faith in their products.
 
Wiring the pump to the mobo is not a guaranteed solution either, because fan control capabilities vary from mobo to mobo. I can tell you that with an ASUS Maximus X Hero, controlling the pump via the mobo is NOT a satisfactory result, mainly because it has limited RPM customization, and you cannot assign the fan curve to GPU temps without hooking up an external temp sensor.
 
Attempting to self-solve shouldn't enter into the equation! EVGA should be doing more to help its customer base. I'm not impressed with the general build quality either, especially with regards to the metal backplate and how it fails to consistently align with the screw holes of the plastic housing. Failures like that are detrimental to the whole experience.
 
I'm supposed to RMA this card, right? Well, easier said than done considering that I don't have a spare card to use in the mean time, and there's no guarantee when I will receive a replacement or if the new card would even alleviate the problem. This is even more concerning given the current climate with COVID-19 forcing many people to work from home.
 
To people who bought the Hybrid card and felt mugged by the experience, at least going from my personal experience, am I going to recommend or buy another EVGA product after this?
 
The answer is no. Not unless EVGA publicly acknowledge the issue, take the initiative and provide recompense.

I was happy with mine before I switched it over hydro.

They do have advanced shopping options where they have a fee and hold a value on card (credit/debit) and you get the card first.

I think you should visit that option if that is a concern. You’re not going to find a video card company that treats you better than EVGA.

Expecting a several hundred dollar item before before acquisition of the damaged product sounds like a terrible business idea.

Nazarus
Another resolution people can’t seem to grasp is, if it’s broke return it and get a new one before you have to deal with warranty.

People let things that “work” sit in their system until it bothers them that it doesn’t fix itself magically.

Patience you must have. -Yoda

Valid points, but EVGA could make this considerably easier for the customer by setting up a replacement programme just for the pump, rather than playing roulette by replacing the affected item with a completely new card. It has been reported that the severity of the pump noise can vary, but it is a marginal difference. The problem still exists.
 
The measured performance of my card is acceptable. Benchmark results are great. I do not wish to replace the card itself, only the cooling solution. If I decided to invest in a NZXT solution for example, everything would be fine and dandy, although I would need to put another $200 down to fix a problem that EVGA caused.
 
It doesn't make sense to RMA both the card and the pump. It only shows that they are complacent with the fact that they have produced a sub-par pump solution that is infamously known for upsetting their customers. Why aren't we being offered a cooling solution where this problem has been addressed?
 
There is official documentation for installing the pump, along with plenty of YouTube videos too, so it isn't a particularly difficult task to replace the pump, especially for those who have built a PC etc.
 
I received a reply from their customer support today, and just as I expected, they did not directly acknowledge the wide-spread issue and were unable to offer any kind of acceptable recompense. They are also wasting my time by asking me to provide a video/sound recording of the issue I am experiencing.

You should not mount your radiator below pump though...

That is not a myth, just a physics: there is could be air in AIO and your radiator take a resevroir role to feed pump with water but not with air. 
So thats why radiator should be mounted higher than pump. If air gets in a pump that will result in horrible sound and shorter lifespawn. 
 
On your video I can't hear air inside pump so maybe you are lucky and there is almost zero air in your AIO, but you should not do that anyway and that is something support could tell you to "fix" and rerun tests.
There is even instruction for that:

 
https://www.youtube.com/watch?v=CV1BT5nMkhg as example what pump running air sounds like.

devilhood
 
Initially, I thought you were talking about whether the fan was in-front or behind the radiator, but I later understood what you were referring to.

That is also not a myth ;)
You are talking about push and pull configurations and it is known that push works slightly better on high rpm while pull works sligthlty better on low rpm, while push-pull configuration works significantly better them both.
The reasons here again physics (aerodynamics): on low rpm push configuration just pushes some air around fan instead of pushing it through radiator (not enough pressure to push it through), while pull configuration on high rpm just sucking some air through gaps between fan and radiator. So it is all about how much air moving through radiator and nothing more.
 

devilhood
Thanks for the response. In this case, the diagram does not match my set up. My pump and radiator are not directly underneath the graphics card. It clears the width of the card completely, and the resulting ambient temperature is also being actively cooled by the airflow from the front of my case.
I understand that this is a recommendation, but it should be considered depending on the type of case you have.
I have placed it in both positions and the outcome did not produce a change in results. My decision to place it at the bottom was considered purely for aesthetic reasons, but I wanted to ensure that the decision would not negatively affect the performance and temperature of the card.
If EVGA ask me to change its position, I would happily do that, and continue to prove to them that their pump is an example of poor engineering.

Well, diagram showing not "radiator under gpu" like you got it literaly, it just showing you radiator above pump level. Again, it is all about air inside AIO loop and that air moving to the highest point of your loop. In your configuration this place is pump itself. It is not about how far away your radiator from gpu - but about fact that it is lower than pump level.
If you place radiator higher than pump - all air will be trapped inside radiator, like this:

While if you place radiator lower than pump or mount radiator with tubes upward - air will be trapped inside pump or near pump intake, in that case pump is moving air and making bubbles inside loop. That affects performance a little (like it is either moving water with bubbles with slightly less thermal capacity or in worse case scenario there is no water inside pump, you are experience somehting called airlock and impeller just spinning with air trapped around not moving any water at all and that will result in instant overheat), but also affect pump noise and pump health because pump uses water as lubricant too. And because you are worried about pump noise - support could tell you to mount your radiator higher than gpu and check noise again. So you can ... "outrun"? ... that suggestion.
But again, I don't hear any bubbles sound in your video, just a "normal" pump hum like I had so it is either almost no air inside your loop or it is trapped somewhere not near impeller or intake tube. In both cases you are lucky guy :) I had large amount of air in both gpu and cpu AIOs and my cpu AIO was running with air for 2 or 3 months because it was mounted at front with tubes at top. Laterly I moved it to top, so there was no air and no bubble sounds plus no intake air restrictions - but pump started dying after 7 or 8 months anyway. **but I switched to custom water at that time so that was not a big deal :)
 
The over concenr is that some air could be trapped inside small bubbles on surfaces like tubing or radiator walls. And this bubbles slowly going up: you need couple of weeks to get rid of all bubbles inside loop. And in your case these bubbles slowly getting inside pump and that is not good for pump lifespawn.
And another "but": if any air will be trapped near impeller in future - you will hear it anyway.

USPS delivered today, so I connected a thermal sensor to the T_SENSOR header on my mobo and can now properly assign the fan curve. It's not directly measuring the core, but the results are close enough. Tempted to take the housing off to see if I can find a more optimal position for the sensor.
 
The issue is now more tolerable.

DeadlyMercury
I just pushed sensor in a hole in backplate above gpu :) That was fine enough to control both pump and rad fans speed. Actual reading was 5C less than gpu temp, but that is enough to understand if card idling or working and ajust fan speed.

Nice, I did the same. After about 5 minutes the backplate temps drop to around 40c and then gradually fall to 34. I've set the pump and fan to turn off at 35 and it seems to work very nicely indeed.
 
Edit: I used industrial grade thermal tape and placed the thermal sensor on the backplate precisely above the core GPU processor and it has closed the gap in temps. Instead of there being an average of about 5-6c, it is now close to being identical, with only a slight delay in how fast the heat is transferred to the backplate.