Why they do not only did like asus did? To change all to high quality ones?

Who says that SP-CAPs aren't high quality? Diversity is good too, you know. ASUS's 60 MLCCs aren't necessarily better. Watch the video Sajin posted above. Derbauer's conclusion is valuable.

Yeah, the MLCC capacitors themselves perform better. No doubt about it. Many smaller capacitors with varying capacities perform better than one large capacitor. I am not arguing that. But just because the capacitors themselves perform better doesn't mean that it will be more desirable for this application if all of the capacitors are the same capacity. Various sizes and capacities are usually preferable, giving better characteristics over a wider range of frequencies; instead of having acute peaks and valleys at various frequencies. Just a general rule. If all 60 MLCCs are the same capacity, the end result MAY be less preferable to pairing 4 x 470 microfarad and 20 x 47 microfarad capacitors. Variety is usually preferable. Either way, in this application, I believe 4 x SP-CAPs plus 20 x MLCC versus 0 x SP-CAPs and 60 x MLCC: it won't make a difference or it will make such a tiny difference that it won't matter.

I disagree. Or, should I say "not necessarily"? I agree, with caveats. And I disagree, with caveats. Therefore I cannot agree with such a simple "it's better" or "it's worse" without testing or discussing the caveats.

4 x 470 microfarads plus 20 x 47 microfarads can easily perform better than 60 x 47 microfarads across a wide range of frequencies.

IF those 60 x MLCCs are a variety of different capacities, that's one thing. But if they are all the same capacitance value, they can be shown to be inferior across a wide frequency range. When dealing with a wide frequency range, variety is often preferable.

The capacitors themselves: better versus worse can easily be proven. When installed on a complex device, its no longer that simple.

If you already understand what bypass capacitors are and why they are used, skip to the 10 minute mark in the following video, and then watch him -- an electrical engineer -- explain why various bypass capacitance values are used in combination with eachother for improved frequency response:
https://youtu.be/BcJ6UdDx1vg

Yea fixed:D You know how? They don't allow to boost over 2000 :)

Crismac1
I agree and if I can't get a card by the time AMD launches and I like what I see I might go that route. I haven't bought an AMD card since 2006. We will see if it's time.

I've also considered this but I'd really like to have ray tracing and DLSS. If AMD has equivalent solutions then it might be my first AMD card.

musiq1086
Is there anyway to know if there is going to be a hardware change to new cards comming out or if they will just not let you clock over 2k anymore 

It's a driver update. They changed they boost table.
No hardware change necessary.

Fighting the disinformation: https://youtu.be/lhyCdraz54s

Ugh, wish this was not an issue

Pilggrimm
Just got a evga 3080 this week and the picture looks like the problematic ones what can I do?

This topic is regarding the FTW3. If you got the XC3 then it is normal to only have 1 group of MLCC capacitors.

Those are also fine.
It was al just driver issues.
eteknix even has a zotac (with 6 of the "bad" ones that is doing around 2100 without any problems)
 
Igor even resolved issues with some psu that couldnt handle the gpu with the old driver
https://www.igorslab.de/en/wonder-how-invidia-the-crashes-of-the-force-rtx-3080-andrtx-3090-will-be-removed-and-still-will-be-removed-even-from-the-power-supplies-analysis/
 
hardware unboxed
https://www.youtube.com/watch?v=lhyCdraz54s
 
Jay
https://www.youtube.com/watch?v=nMns8t4OhI0
 
nvidia statement
https://www.nvidia.com/en-us/geforce/forums/geforce-graphics-cards/5/400499/rtx-3080-board-stability-new-driver-capacitors/
 
It was all way way way overblown so the witch hunt can be stopped now.

All I need to know is do I get my card replaced for the ftw one since my current card is crashing on most games..? Even with the latest driver. Dont use words like overblown if something is a non-issue for you... put yourself in the shoes of the person having spend so much money and then have to deal with a product not working as it should.

Pilggrimm
All I need to know is do I get my card replaced for the ftw one since my current card is crashing on most games..? Even with the latest driver. Dont use words like overblown if something is a non-issue for you... put yourself in the shoes of the person having spend so much money and then have to deal with a product not working as it should.

If you bought a new 3080 and it's crashing on you then you need to open a support ticket so EVGA's techs can help you out. As Jacob's original message says, EVGA didn't ship any cards with the problematic cap configuration. 
 
You should also update your drivers to 456.55 if you haven't done so already, those fixed a lot of the crashing problems for people.

Pilggrimm
I don't overclock anything. Cpu, mem, gpu all at factory. I'm running the latest driver. I did open a ticket.. someone said I can also check the forums which is what I'm doing here.

What happens with the crashes? Does it go to the desktop? Does your pc reboot? Do you get a bsod? Did you check in gpu-z what happens before the crash?
What psu you got? Is it connected with 1 or 2 different 8 pins?
Did you try to delete everything with ddu and reinstalled the drivers?
Can you test it in another pc or another psu?
 
We wanna help but we need more info m8.
 

Axejess
Igor even resolved issues with some psu that couldnt handle the gpu with the old driver
https://www.igorslab.de/en/wonder-how-invidia-the-crashes-of-the-force-rtx-3080-andrtx-3090-will-be-removed-and-still-will-be-removed-even-from-the-power-supplies-analysis/
 
 
... 
It was all way way way overblown so the witch hunt can be stopped now.

Not quite. Igor sums it up nicely in the conclusions on page 2. Paraphrasing, he asks why this wasn’t known earlier. And Igor is right. This was a rushed launch and Nvidia did not give AIB manufacturers the time or tools needed to properly design and provision these boards. It caused needless expense and delays for the rework and band aid patches and fixes to get working cards. The driver patch reduced performance but was necessary to prevent a total meltdown of the product launch. Slightly, maybe imperceptibly for some in some applications, but it still reduced performance to kludge for a hardware issue.
 
Without this noise issue, we might even be seeing many more cards hitting the 2.1GHz clocks Nvidia originally was shooting for. It’s spilled milk, but if we had those clocks, that’s a significant increase in performance. Bet the “Ti” cards hit 2.1GHz.
 
The delays and shortages have kept cards out of the market and gave scalpers a footing to make bank. Had Nvidia done things right, the time lost to proper testing and design would at least have been partially made up by not having to hold cards off the market, maybe fewer scalpers from less incentive, etc.
 
I wish Nvidia learned a lesson but I doubt it. Nvidia buggered this launch and I bet they do it again next generation. It’s a crappy attitude company that is able to coast on otherwise good technology. I hope they get hit by a class action lawsuit. Seems money is the only thing companies understand these days. They opened themselves up to that possibility when they took performance back to fix their mistakes.

Frammish

Axejess
Igor even resolved issues with some psu that couldnt handle the gpu with the old driver
https://www.igorslab.de/en/wonder-how-invidia-the-crashes-of-the-force-rtx-3080-andrtx-3090-will-be-removed-and-still-will-be-removed-even-from-the-power-supplies-analysis/
 
 
... 
It was all way way way overblown so the witch hunt can be stopped now.

Not quite. Igor sums it up nicely in the conclusions on page 2. Paraphrasing, he asks why this wasn’t known earlier. And Igor is right. This was a rushed launch and Nvidia did not give AIB manufacturers the time or tools needed to properly design and provision these boards. It caused needless expense and delays for the rework and band aid patches and fixes to get working cards. The driver patch reduced performance but was necessary to prevent a total meltdown of the product launch. Slightly, maybe imperceptibly for some in some applications, but it still reduced performance to kludge for a hardware issue.
 
Without this noise issue, we might even be seeing many more cards hitting the 2.1GHz clocks Nvidia originally was shooting for. It’s spilled milk, but if we had those clocks, that’s a significant increase in performance. Bet the “Ti” cards hit 2.1GHz.
 
The delays and shortages have kept cards out of the market and gave scalpers a footing to make bank. Had Nvidia done things right, the time lost to proper testing and design would at least have been partially made up by not having to hold cards off the market, maybe fewer scalpers from less incentive, etc.
 
I wish Nvidia learned a lesson but I doubt it. Nvidia buggered this launch and I bet they do it again next generation. It’s a crappy attitude company that is able to coast on otherwise good technology. I hope they get hit by a class action lawsuit. Seems money is the only thing companies understand these days. They opened themselves up to that possibility when they took performance back to fix their mistakes.

Yeah they maybe rushed it and i do agree aib's should have had more tools or better drivers to test with like gamernexus also pointed out. And build up more stock.
But I don't agree that if that would be done that everyone would have had 2.1 chips. Maybe on tsmc sure. And if you look at benchmarks there don't seems to be that much dif between 1900-2050ish cards anyway so I don't think it would have matter that much scaling wise. Higher is better sure but that would cost more power watt wise and the only thing to reduce that would have been tsmc. We don't know the whole story but I guess that it also has to do with that tsmc is quite full order wise with amd/consoles/ryzen/new navi cards and there other customers wich also maybe why nvidia unfortunate had to go with samsung this time. Who know maybe we will hear the real story in the future what the real reasons where beside this. We can only speculate that's all.
 
And the benchmarking I don't see that much of a difference between the drivers it's not like there is so much gimping that it cost like 5%. I only saw like 1% and sometimes 0 or a little + even. I don't think it matters that much. Some people even are reporting higher clocks now on there cards and get a improvement instead.
 
cap noise yeah well I also think it has to do with the rest of the card. If you look at der8auer who removed some caps and replaced them with mlcc he got a stagering 15mhz more. https://www.youtube.com/watch?v=ud6NrbJllzkSo that's not that much of a difference is there? I don't see how that would let every card hit 2.1+
I still think nvidia is right with there statement that it depens on much more factors. Evga seems to have 270's caps on the back while others have 470's. Why? I trust them that there is a reason for that maybe the other parts of the board have much more and that's why they don't need bigger ones? Maybe bigger ones has more downsides for them? Who knows I sure don't.
 
Scalpers also has to do with the times they also do this with things like iphones and concert tickets ect. That just have to do with the modern times as well people want to earn money that's the unfortunately truth of capitalism. There are even sites now where you can buy your own bot(s) for that. Yes it's annoying but that's the modern times we live in. Best way would be if you have the patience wait till all the dust settles and if you can see what the other camp brings on the table. I myself am locked on nividia cause of my c9 and my gsync module monitor. Just remember it's only a videocard there are much more things in the world atm (corona).