demowhcc
I just want to confirm that if flashing to the X3C bios, all readouts from GPU-Z remain correct? IE- clocks, voltage, power etc?
 
Or does the altered 2x8 pin bios mess with readings?

Clocks and voltage are correct everything else related to power isnt.
as mentioned here : overclock.net/threads/evga-3090-ftw-power-limit-bypass.1774611/#post-28671577

It would be nice if EVGA would just respond and tell us if they are working on a solution or if they don't have a solution. Any answer would be fine.

Nereus

cerealkeller

Nereus
So his conclusion is that the 500W bios doesn't work (or makes no difference from the 450W one) because the PCIe connection on the 3090 FTW card has a 75-80W limit that can't be fixed without a shunt mod, correct? Maybe I missed it because the guy talks too much, but what limit do all the other cards have on the PCIe slot? Do they have any limit? He does a shunt mod that ups it to 120W (I think that's what he said) but repeatedly said to do it at your own risk.. so by extension, do all other cards have an inherent risk now by not having a 75-80W limit there? Or is that an artificially imposed limit by EVGA, in which case I again wonder what the other cards do there... sorry if he answered this and I missed it, it was late, I was tired, and again.. dude talks too much lol.

PCI-E devices on a x16 interface shouldn’t be exceeding 75w. 75w is the power limit for a PCI-E slot. Just like 6 pins are 75w and 8 pins are 150w. You can draw more power from them if the card calls for more power, the power supply will oblige until something breaks, or there’s a hard cut from software or some other limiting factor. There’s a 10 amp fuse on the PCI-E slots shunt (this is the only way I know to explain it) and there’s a 20 amp fuse for each 8 pin. I’ve read that as long as you don’t exceed 120 watts for the PCIE and 240 watts for the 8 pins, you shouldn’t blow a fuse. Someone here has mentioned that a card did blow a 10A fuse, I think that’s the case, where someone had an 8 mOhm shunt mod on all the shunts. Obviously 8 mOhm is too much and exceeded the120w limit.
Other manufacturers have lower power limits set for the PCIE slot. I don’t know the numbers off hand, but I remember seeing the FE was 66 watts and some others are 50w. With three 8 pins, there’s no need to draw 75+ watts from the PCIE slot, Even at 500w.
What I have observed is the card cutting power because it’s drawing too much power from the PCIE slot even when the 8 pins aren’t maxed out. It’s really bad design.

Thank you for sharing that knowledge. Is this something that can be adjusted by bios/firmware update, or is this strictly a hardware issue? I thought the reason for having 3 x 8 pins was to handle the higher power of this card instead of overloading the PCIe slot... so this is a load balancing issue? Again, not an electrical engineer here. I may have already asked this somewhere before..
 

That’s the million $ question isn’t it? If it were as simple as changing a few values in the BIOS, wouldn’t they do it? A month ago I would have said yes, but at this point my perception of EVGAs customer service has been brought into question. Which is ridiculous. A simple reply by someone who is affiliated with the company would satisfy me right now. It’s so weird how they’re dead silent and for the fact it’s been like a month of silence.
Edit:
Apparently Jacob chimed in like two weeks ago. I never saw it since it was before I joined this thread

The XOC bios have a 420w target and can peak at max 500w.
Myself i saw over 510w when i play games but that bios is not intended to continously be at 500w.
Its depend of the game and most of the time its stand at 440w-475w.
 
Not coz you try some benchmark you have to see 500w, its depend of the workload.
 
The only thing it can be good its increase a bit the power target at 445w so we can get +25w on average.
But i dont know by doing this it can bring more crash.

Akriant
Well. With the 500w bios and looking at live feed via Nvidia OSD (which works better than EVGA's, which is sad) I saw my card going up to 495w. My OC stability goes out of the window past +80 on core though. Idk if it's temperature related or not, but, I've just splurged a stupid amount of money on the optimus waterblock (just saw it on sale today for like an hour lol), so I guess I'll see how this card performs paired with 480mm rad. 

Ahh that was probably thanks to me! I just canceled my order and they just responded about 3 hours ago confirming it was canceled and processed the refund. Friendly people, no problems canceling, and I also asked to buy an o-ring for my V2 block and they are sending one for free! I would really love to have that GPU block, but it's just too much $ when I thought about it, and plus I've had a couple of brief black screens so I'm a little hesitant to spend that much more money on this card now. I'll try for an EK block or maybe EVGA HC if they sell them.
 

Akriant
Well. With the 500w bios and looking at live feed via Nvidia OSD (which works better than EVGA's, which is sad) I saw my card going up to 495w. My OC stability goes out of the window past +80 on core though. Idk if it's temperature related or not, but, I've just splurged a stupid amount of money on the optimus waterblock (just saw it on sale today for like an hour lol), so I guess I'll see how this card performs paired with 480mm rad. 

It's evident with a waterblock the card will boost at least 2100mhz and i dont think it will drop and if you ad +85 maybe you will keep 2150mhz without crashing. 
Me on air when i start gaming my card boost at 2100mhz then drop to 1990mhz, iam sure its related to temperature. I put +110 and + 900 memory for gaming.
I wanna get the just the hydro cpper block but i dont know if they will sell this. 

Nereus

cerealkeller

Nereus
So his conclusion is that the 500W bios doesn't work (or makes no difference from the 450W one) because the PCIe connection on the 3090 FTW card has a 75-80W limit that can't be fixed without a shunt mod, correct? Maybe I missed it because the guy talks too much, but what limit do all the other cards have on the PCIe slot? Do they have any limit? He does a shunt mod that ups it to 120W (I think that's what he said) but repeatedly said to do it at your own risk.. so by extension, do all other cards have an inherent risk now by not having a 75-80W limit there? Or is that an artificially imposed limit by EVGA, in which case I again wonder what the other cards do there... sorry if he answered this and I missed it, it was late, I was tired, and again.. dude talks too much lol.

PCI-E devices on a x16 interface shouldn’t be exceeding 75w. 75w is the power limit for a PCI-E slot. Just like 6 pins are 75w and 8 pins are 150w. You can draw more power from them if the card calls for more power, the power supply will oblige until something breaks, or there’s a hard cut from software or some other limiting factor. There’s a 10 amp fuse on the PCI-E slots shunt (this is the only way I know to explain it) and there’s a 20 amp fuse for each 8 pin. I’ve read that as long as you don’t exceed 120 watts for the PCIE and 240 watts for the 8 pins, you shouldn’t blow a fuse. Someone here has mentioned that a card did blow a 10A fuse, I think that’s the case, where someone had an 8 mOhm shunt mod on all the shunts. Obviously 8 mOhm is too much and exceeded the120w limit.
Other manufacturers have lower power limits set for the PCIE slot. I don’t know the numbers off hand, but I remember seeing the FE was 66 watts and some others are 50w. With three 8 pins, there’s no need to draw 75+ watts from the PCIE slot, Even at 500w.
What I have observed is the card cutting power because it’s drawing too much power from the PCIE slot even when the 8 pins aren’t maxed out. It’s really bad design.

Thank you for sharing that knowledge. Is this something that can be adjusted by bios/firmware update, or is this strictly a hardware issue? I thought the reason for having 3 x 8 pins was to handle the higher power of this card instead of overloading the PCIe slot... so this is a load balancing issue? Again, not an electrical engineer here. I may have already asked this somewhere before..
 

Not a load balancing issue. Something else, that no one can competantly confirm.

The 490w avg i sustained drew 82-85w from the pcie slot

Turbo-12R

cerealkeller
Unfortunately his channel is extremely tiny, an insect in the grand scheme. But hopefully it helps get the ball rolling.
Even my stock BIOS was messed up, drawing 80+ watts from the PCI-E slot.

There's quite possibly a reason for that.  He's hard to stomach and listen to.  As a middle aged guy, maybe I'm a bit too old school, but the vocabulary he uses is very unnecessary and not needed to be able to get a point across.  I labeled him correctly here in a previous post and hurt someone's feelings here for doing so, but what I said was accurate.
 
Maybe this guy has a beef in general for EVGA, maybe they did not send him a sample to review like they did to some other reviewing outlets and it upset him...who knows.  

you realize he started youtube 3 months ago right... he is growing incredibly quickly for the techtuber space,
 
he's definitely a hot head but he actually knows way more **** than like 99% of people,
 
he has an insane amount of supporters funeling him money and is already getting review samples from EK with 2k subs

sirien

Turbo-12R

cerealkeller
Unfortunately his channel is extremely tiny, an insect in the grand scheme. But hopefully it helps get the ball rolling.
Even my stock BIOS was messed up, drawing 80+ watts from the PCI-E slot.

There's quite possibly a reason for that.  He's hard to stomach and listen to.  As a middle aged guy, maybe I'm a bit too old school, but the vocabulary he uses is very unnecessary and not needed to be able to get a point across.  I labeled him correctly here in a previous post and hurt someone's feelings here for doing so, but what I said was accurate.
 
Maybe this guy has a beef in general for EVGA, maybe they did not send him a sample to review like they did to some other reviewing outlets and it upset him...who knows.  

you realize he started youtube 3 months ago right... he is growing incredibly quickly for the techtuber space,
 
he's definitely a hot head but he actually knows way more **** than like 99% of people,
 
he has an insane amount of supporters funeling him money and is already getting review samples from EK with 2k subs

 
I just deleted a lengthy response to this.  It was not worth the argument.  
 
 
By all means, feel free to follow him.  

Good evening,
i got a question, if i flash the card with the xc3 bios

2x8 pin 400 watt

do i get more power than 450 ?
soo 3 x 150 and 1x 100 watt pci-e slot ?

is this save to draw 550 ish watt?

PadinnPlays
If I'm understanding recent posts correctly, the XC Ultra bios will allow the card to draw 500w? That points to a bios bug and not an issue with the hardware. Am I missing something?

 
You are.  You obviously missed the part where you have to shunt mod the card, while simultaneously rubbing your stomach, patting your head, and turning in a circle three times.  Make sure to nibble all your fingernails down to the quick too prior to doing this.

Well I've seen a lot of complains for me the XOC BIOS for my FTW3 works very good, better than the stock ones, still is limited YES, but I can tell that can hold these clocks steady in games and will drop to 2025 - 2055 but mostly stays solid on 2070. As for the memory it's a beast and I'm still on Air with no mods at all. On benchmarks will go 1995 and jump to 2010 and over, on some ocasions when the benchmark is heavy will do 1950 and go above to 1985 and then 2010 etc. I got pretty nice scores on HWBOT for a Stock Card, and for Gaming it's awesome. Hope they release a Waterblock soon as I have to pump the fans and put another cooling the backplate, specially the caps zone that heats a lot. I tried undervolting it and works very nice also, of course less clocks but always above stock :)
 
Here's my game settings:

 

bavor

arcky
*crickets*
https://youtu.be/Lmd48SVbXso
*crickets*

EVGA has about 70 total full time employees. They aren't Asus and MSI with thousands of employees.  Its not like EVGA can spare engineers to work full time on a BETA BIOS power draw issue when they still are working on new soon to be released products with a limited time frame between Nvidia releasing info to the manufacturers and the product release dates.

it's not only that, looking at buildzoids latest video about the components on the ftw3, I pay premium price for a card, but it does not have premium components on it, rather mediocre.

bavor

arcky
*crickets*
https://youtu.be/Lmd48SVbXso
*crickets*

EVGA has about 70 total full time employees. They aren't Asus and MSI with thousands of employees.  Its not like EVGA can spare engineers to work full time on a BETA BIOS power draw issue when they still are working on new soon to be released products with a limited time frame between Nvidia releasing info to the manufacturers and the product release dates.

Hey man... you and others sit here and post like we don't have any options.  Like "how dare we" question evga on their artificially limited cards!?!?! 

Let me clue you in on how this works. EVGA burns its most demanding enthusiasts who are spending at the HIGHEST LEVELS... then goes full radio silent for a MONTH about it... and we will bail on them and spend AT THE HIGHEST LEVELS with some other company. Simple. ASUS and others, would love to have our cash. EVGA could have stopped this any time they wanted by being present. Instead, they want to treat us like mushrooms. Well, that has consequences. 

We paid TOP MONEY for basically a reference card. They clowned us. If I'd have paid $1400, I wouldn't be upset.  

arestavo
Well.

Bye.

Oh you're super clever. Thats fine. You and EVGA can marginalize me all you want. I'll do what I do.