You need to reinstall X1 after rebooting on the normal bios.  I had this problem too.
 
You can verify the card is on the normal bios by looking at GPU-Z, advanced tab, NVIDIA BIOS dropdown, and see the max power limit is 450 instead of 500.
 

dexmex88

arestavo

dexmex88

arestavo

dexmex88
Same issue as everyone else, no increase in draw. Weirder issue, flipping the BIOS switch back... I still am on this BIOS. Anyone else have that issue?

Requires a reboot.

Yeah I've rebooted multiple times and went back and forth between the 2 BIOS's... .

Sounds like you flashed both slots then. Or perhaps you need a full shut down? Maybe something weird going on with Windows.

I'm not sure how I could have flashed both slots considering I was on the one slot the entire time and didn't change over until I figured out this wasn't working.

Just running some test. Most watts I could get was 479 a lot above 470 when cards up above 90 % .My card being clock down by temp. Wish I had a water block to test new bios.

so there have been people who tried to flash Asus 480w bios on to their ftw3 cards and have had same issue.
In fact, they saw a reduction in performance since it changed the behavior of the pcie being capped at 75w (vs 80w you get with the 450w bios) so they actually saw a reduction of 5watts by using the 480 strix bios
 
So im under the impression this is not fixable with software and its an intentional hard limitation set in the circuitry by Evga. 
Maybe it's to segment their kingpin card? just speculation 

Guys... this is really simple.. If there is even ONE ftw ultra 3090 getting over 450 (Which many have claimed) - Then we all can. I'm pretty sure some people had different firmwares or came from a different bios. EVGA won't let us down. After all, we need the frames . FOR THE WIN!

 

god503
The max hit 115% power and 485 watts. used precision X set at 119 power, 100 voltage, +175 on the GPU and +600 memory.  this is on a threadripper 3 24 core, thiachi mobo and corsair 1600 power supply.   I am fully convinced that the air cooling cannot handle the 119 power level.
 

did you do any benchmarking before to compare to now? I want to know that its truely working. Sometime the board power draw goes over the limit so even before this bios its possible to draw over 450w on the vanilla bios at times before it drops back down. I'm trying to gauge if its actually working for you

I’m on an open air test bench, fans at 100% and not going over 54c...yet can’t get the board to draw more than 425w average, 450w max across all benchmarks and games. It’s not temperature. Something is not working. Would be great for a response from EVGA...

dam_j
 
But didn´t you guys know the power limit/target before you buy ? Did you all just blindly buy and be now disappointed that EVGA doesn´t have a greater target as shipped ?
 
I don´t buy a car with 200mp/h and afterwards give the manufacturer a hard time because it can´t do 250mp/h...

I actually think your car analogy is a great one. What has happened is akin to Tesla pushing out an optional update to their customers that promises to increase the horsepower of their vehicle, only for the customers to find out that in order to utilize the update they would have to alter their vehicle in a way that voids their warranty, shunt modding in the case of the GPU. And through all this Tesla (or EVGA) has remained silent and not even acknowledged the issue, or even provided information to its customers as to whether or not there is an issue or if this is working as intended.

I don't think *anyone* would be upset if EVGA had released this BIOS update w/ a disclaimer that stated outright this BIOS will not function as intended without additional steps being taken by the end-user that *would* end up voiding their warranty. I also don't think *anyone* would be upset to find out that their GPU doesn't perform any better at 500 watts versus 450 watts because that's silicon quality for you. But at this point, the *vast* majority of users can't even get the card to pull over 450 watts while other AIB cards *on the same BIOS* are doing so w/ no issue and reaping the benefits of the increased power limit.

Seems like (at least at this point) this BIOS was a play to increase the perceived value of the card to prospective enthusiast buyers who only read a short article about the release of the BIOS, but don't actually do any research into whether or not the BIOS actually does what it says it does. They now see "EVGA 3090 FTW3 Ultra w/ 500 watts PL" and compare it to "Asus Rog Strix 3090 w/ 480 watts PL" as opposed to the less favorable comparison of "EVGA 3090 FTW3 Ultra w/ 450 watts PL".  

The greatest irony of it all is had EVGA not even bothered releasing this BIOS this would be a non-issue. EVGA 3090 FTW3 Ultra buyers (like myself) would be happy w/ our 450 watts. But when you commit to something as a company you need to follow through, and this BIOS, in many cases, is not doing that.  

GTANY
god503 : What games/benchamrks did you use ? Was 485 W only a peak or a constant value during many seconds ?

Thats what im thinking. The way its working for the other aib partner cards, they are really pushing around the 500w mark and sustaining it for large periods. 
If it's truely working you will notice a good improvement in performance it wont be trivial 

Here’s my theory: The fact that some are peaking around 480/490w is not evidence that these bios work. It’s evidence that the pcie slot is BRIEFLY peaking close to 85w which allows the pci pins to BRIEFLY pull more power according to this locked ratio. But most of the time the slot is at it’s rated 75w pull, and the pins are locked to 125w according to the fixed 1:1.67 ratio at this rate, and here we are...

The bios are installed, gpuz and px1 show the 119%/500w limit. I’ve never had my card on the normal bios switch since around 5 seconds after I removed it from its box. There’s a problem with either the implementation of the bios or the cards themselves.

God: you said your maxed temps were 81c...mine are 54c...and you’re claiming your bios are working and mine isn’t due to cooling? What?

I would reflash the bios and make sure your on OC mode.  


Tried that one already as well as both oc and normal xocs, still max 450 in use, 500w in theory.

I will attach the ezplug and also ensure 3 separate 8 pins to my card tomorrow.

Today I use 1 8pin, 1 split (2) 8pin and no ez molex 4pin plug. I will update you on it after I tried.

thebc2

dam_j

thebc2

dam_j
I´m a little bit shocked about the behaviour of some people here over a BETA bios that´s not nearly a day out...

I think there are a few things causing anxiety for FTW3 Ultra owners.
 
1) No acknowledgement that there is any type of issue or that it's being looked into.
2) This BIOS apparently working just fine (in regards to upping the power limit) for competing (and much cheaper) 3x8ping 3090 cards.  Also, if 450W is the true limit of this board, thats also the power limit of their 3080 FTW3 Ultra, huge kick in the gut - double the price for the same power limit?
 
Many of us paid up for the FTW3 Ultra expecting a card that would be a decent overclocker with increased power limits.  Otherwise I would have bought a Zotac or even an XC3.  I don't see the value prop of this card if it is in fact gimped.  And until we hear otherwise from EVGA, all we are left with is the community's work trying to assess the situation with this BIOS.

But didn´t you guys know the power limit/target before you buy ? Did you all just blindly buy and be now disappointed that EVGA doesn´t have a greater target as shipped ?
 
I don´t buy a car with 200mp/h and afterwards give the manufacturer a hard time because it can´t do 250mp/h...

I don't know if this is your first ever round of graphics card releases, but typically cards are tiered based on the amount of overclocking head room they have.  BIOSs are frequently released later that sometimes up those power limits.  Releasing a higher tiered product already at it's physical power limit (aka no head room) would be a huge oversight, especially for a card that costs 1799 and is billed as an overclocking variant.
 
Anyway, continue with your trolling.

 
 
Actually dude you are wrong. Cards and not tiered based on the amount of overclocking headroom they have. That would require extensive binning by EVGA and the time it would take would not be worth it for them to spend. What they do is find out if a card will hit certain clocks and sell them as factory overclocked. And card from any tier will always be able to have more or less headroom than any other. It's called the silicon lottery. A "gaming" variant might overclock higher than an "ultra" card. They aren't testing for top end clocks or headroom. They are only making sure a chip can hit advertised clocks. There is essentially no reason to ever buy a higher end card if you are manually overclocking. You have the same shot with any card. The only difference this time is that FTW card have a 3rd connector so you MIGHT be able to get more out of one vs an XC3 simply because these Ampere chips are so power hungry. EVGA segregates these cards so they have a reason to charge more. It's all marketing BS if you are going to OC manually. The "tiers" are so people who don't OC manually have a reason to spend more. Why do you think Ultra cards have been so much more available so far? Not because they are all golden chips, it's because most chips will hit the meager boost clocks they offer and they can charge more for them. And at this point people will spend the extra money just to get any card. I'd buy a gaming FTW3 over an ultra FTW3 in a heartbeat if they were both available to buy. It's throwing away $80.

Did anyone try the strix bios on ftw3 ? If so did that go to 480w easily ?

justin_43

zogge
Did anyone try the strix bios on ftw3 ? If so did that go to 480w easily ?

Now that's a good question. Then you would know if it's a hardware limitation on the EVGA or BIOS limitation.

 
I believe someone has (its in this thread somewhere) with the same result as the EVGA provided bios which is why it was pointing to a card limitation.