Any ideas on what is going on with the 8-Pin #3 Power draw on my new 3090 FTW3? These were the max numbers on my first Port Royal run on my card. Default settings (no OC), running the OC Bios (same results with the Normal Bios). Power Supply is a 1300 G2. Please let me know if I missed any pertinent info that you guys need.

Thanks!
 
Lester

There is no official fix.

bmwm3smg
Well, as far as I understand everything, I'd much rather have one of my GPU power cables pulling an extra 8 watts, as opposed to my PCIE slot overdrawing. At least this way if something fries, I think it will be limited to the 8-pin power cable and possibly the PSU, not the motherboard... I'm not sure if the GPU would be damaged if the 8-pin fried, anybody know?

Thanks guys! :)

I think you are missing the point of the problem.  The card is limiting itself to lower performance because 8-pin #3 is *only* pulling 75w when it should be pulling 150w.  

I don't get it. All these 3090 FTW3 card use identical design and components. Why is there such a  difference in plug power usage between cards even when the power supply is not the issue. Bad design? Bad components? 

What is the Problem?

Feklar
I don't get it. All these 3090 FTW3 card use identical design and components. Why is there such a  difference in plug power usage between cards even when the power supply is not the issue. Bad design? Bad components? 

EVGA used a downgraded inexpensive analog voltage controller instead of a commonly-used more expensive digital voltage controller.  When they did that, it is possible that their implementation became more heavily reliant on resistor, capacitor, and or/inductor values for proper analog measurement; and they may have not adequately taken into account manufacturing tolerances for those supplementary components.  Additionally, if they cheapened out on the voltage controller, what's to say that they didn't decide to use cheaper supplementary components with looser manufacturing tolerances as well?  There are different standardized tiers in manufacturing tolerances, and you have to pay different prices for a higher tier component with better guaranteed tolerance.  Standard tolerances range  from +/- 20% of rated value for cheap components, to +/- 1% of rated value for expensive components.
Manufacturing tolerances -- and not adequately engineering with manufacturing tolerances in mind -- can easily explain why different cards read different values and some are more unbalanced than others.
 

mberohn
GPU PCIe +3.3V Input Power (est) is also holding firm at 0.336W. Is this too low, or is the HWiNFO64 showing the decimal in the wrong place?

As far as I know, these cards don't use the +3.3v power input, so I would expect the value indicated would be very low, and would simply be a measurement of error.

Turbo-12R
To those who are having the issue, have you tried swapping the placement of the PSU cables to see if its always on #3?

Yes, they have.  The problem cards have the problem consistently until the card is either replaced or an unofficial BIOS is used.  I understand blaming users and systems, but, no offense, too much blame is being put on users.  I don't understand why users are being blamed when there is a mountain of evidence to the contrary.

As far as I can tell, there is no problem. Each GPU is acting similarly with similar power draw percentage differences between the inputs. The BIOS controls the power draw of each input. Unless you test the GPU's in pretty much identical systems with the exact same settings and the exact same game or bench, you cannot tell anything. And of course, software readings are notoriously unreliable for much beyond 'idiot lights' for most part. Whatever software you're using is dependent upon the accuracy of the sensor it is supposedly reporting. Is this actually a design flaw or as intended?
Is what others are looking for is that both PCIe 8 pin are more drawing more equally? I seem to recall something about the design of the PWM where one input is treated more like a 75 watt 6 pin than a true 150 watt 8 pin input.

ty_ger07
 
Read the part of your quote I was replying to.  It's not a cable or power supply issue.

Regardless, I'm not sure where you get the "blaming" part.  I was asking a simple question.  Additionally, do you even have one of these cards?  Your profile suggests that you do not.

I'm glad you were here to get me back on track as to what my intentions were.  Thank you.
 
Yet mine and others work as intended.  All the while, you continue to be a naysayer, comment off of hearsay and speculation.  Ok...got it. 

Because I am not jumping on the "EVGA 3090's are a bunch of junk" bandwagon, somehow I am belittling people?  Heh...quite the stretch there buddy.  I think it's more inversely that you are getting irritated with me because I do not have any issues.  I'm just here saying that not "ALL" of us are having these issues.  Get it?

*eats popcorn whilst enjoying a 3090 that knows how to power balance properly* 
 
By the way, this is not an issue that effects "everyone" - but a lot of 3090s do have this problem in the FTW3 variety of the card.  As indicated in reviews of the card, the FTW3 of this generation (both the 3080 and the 3090) is the closest it has every been to a "Reference" model - in fact most would say in a lot of ways the FE is actually a better card (which it is, sort of) - and it is definitely not worth the price paid.  I am saying this as someone who had a 3090 FTW3 Hybrid that had the power limit issue of not being able to hit anywhere near "500W" on the supposed "500W" XOC Beta BIOS - which maybe, someday, they'll fix, maybe they won't... one would assume so, but by the time it's "fixed" the 3090 FTW3's reputation is already so FUBARed, I doubt it will matter.
 
For anyone looking at getting a 3090, I'd highly recommend if you're willing to spend the $1799+ on the FTW3, that you give serious consideration to queuing for a Kingpin instead.  If you can fit the AIO in your system, it is well worth the extra bit of money required to get one of those instead.   I would say as far as 3090s go, the FTW3 (Hybrid or regular air cooled Ultra) is not worth the asking price, not that any 3090 is, let's be real, they're overpriced period, but the FTW3 especially is not worth the mark up.  The Kingpin however, is an upgraded PCB, VRMs, all of that good stuff - and it's cooling is just plain better, both stock with the AIO, and if you're willing to invest another $60 or so in some heatsinks & thermal pads, it runs so much cooler than regular 3090s, it's kind of ridiculous.
That is one of the major drawbacks on the FE, from what I hear, the stock memory cooling is not great.

Turbo-12R
I think it's more inversely that you are getting irritated with me because I do not have any issues.

No, I got irritated with you (and others like you) for:
1) feeling like you have to come here and tell people that you aren't having these same problems -- as if somehow that helps the person who is having problems -- when you could have instead posted nothing at all and been equally helpful
2) implying that the person is somehow doing something wrong, has an inferior system, or is unable to effectively perform troubleshooting; when there is an overwhelming amount of evidence to show that this is not a rare problem and there is no officially-supported method to fix this problem
and 3) arguing about how common -- or not common -- this problem is, as if there is data to back up either argument; and as if it really is helpful -- either way -- to the person who is having a problem, to argue about how common or uncommon the problem is.
 
 
And just like that Badboy64 is here, like usual, to provide the same unhelpful comment.
 
Edit:
It is in poor taste.  It is unhelpful.  It lacks empathy.  And it is often condescending.  It comes off as 'I am not seeing this problem with my card, so you must be doing something wrong'.
There are threads in this forum where users have posted about how much they like their video cards and how they haven't had any issues.  Why don't you focus your efforts empathizing with them about your shared positive experiences?  Let others empathize with these people's negative experiences and talk facts about those experiences, instead of throwing the thread off-subject into an argument about how common or uncommon the issue is; an argument which does nothing to help the user with their issue and which can never be won, by either side, because EVGA is a privately-owned company with no published sales data and no shareholders to address.

Are you guys using 3 separate pcie plugs or one daisy chained?
Mine seem to be drawing hair more out of the third PCIE connector compared to #1 and #2
At 481w max draw
PCIE#1: 131.4
PCIE#2: 133.8
PCIE#3: 144.1
Based on the GPU-Z graph, it looks like #1 and #2 were rock solid in terms consistent power draw. #3 was the one that was moving up and down but not as dramatic as what folks are reporting above