TechBuyingHavocI don't know about the others but I have ONLY bought AMD cards (I have a 6700 XT right now, I am adversely affected) and I am complaining because this is a worse consumer experience. I will complain about this regardless of the company. In this thread, it is AMD. When it is Nvidia's turn, I am always there to raise a stink.

It looks a lot like some AMD defenders are trying to pigeon-hole AMD complainers as Nvidia-only consumers to reduce the legitimacy of their statements. Attack the messenger and not the message, right? (See what I did here?)

I'm not referring to you specifically, I directed my comment at the Nvidia only buyers who are still beating the dead horse, the same people who always love jumping at the chance to yell "AMD bad" and the "but mah Nvidia marketshare".
How exactly are RDNA2 users affected right now? Are any games refusing to run with the current drivers? What makes the consumer experience worse not getting a new driver every week? The majority of AMD users have moved on as driver updates didn't end and gpu's aren't being turned into bricks.

And no, IMO the Nvidia only buyers have no legitimate argument here, because they never have any intention in buying an AMD card, Nvidia only buyers for some weird reason have concern for AMD's drivers yet never called out Nvidia for having buggy drivers, black screens or crashing, and like I said those same people have shielded the jacket man for bad or misleading PR as being ok because selling more products apparently makes it fine to potentially screw over users.

Hecate91I'm not referring to you specifically, I directed my comment at the Nvidia only buyers who are still beating the dead horse, the same people who always love jumping at the chance to yell "AMD bad" and the "but mah Nvidia marketshare".
How exactly are RDNA2 users affected right now? Are any games refusing to run with the current drivers? What makes the consumer experience worse not getting a new driver every week? The majority of AMD users have moved on as driver updates didn't end and gpu's aren't being turned into bricks.

And no, IMO the Nvidia only buyers have no legitimate argument here, because they never have any intention in buying an AMD card, and like I said those same people have shielded the jacket man for bad or misleading PR as being ok because selling more products apparently makes it fine to potentially screw over users.

Right, but when you're defending EVGA because "big bad meanie Nvidia killed it", you're allowed to say your piece and others are untrustworthy

Mate, pull your head out of the sand, you really think an AMD GPU is a foreign object to any of us here?

Dr. DroYou're confounding "25.10.x" as in, "2025 October" public release and "25.10 code base".

Ah, I get what you mean. But still, the question remains: why do you think this is a problem, or will be a problem in the future?

All in all, I'm feeling some malice behind these "just wait, old drivers will blow into your face eventually" fortune-tellings.

Super XPNo gaming optimizations, no big deal, it's not like games are going to run bad on these older graphics cards right?
It's good that AMD separated the old GPUs with the modern ones to better concentrate on each Gen for best software.

Finally, someone who gets it! :)

RJARRRPCGPI went on the Arc train due to this.

I wouldn't trust Intel to keep arc going, given that they've taken an investment from Nvidia.

RJARRRPCGPIt's one thing to make Navi 1x "legacy", but making Navi 2x "legacy" or legacy-like? WTF! Navi 1x never had RT, so it does look like time to retire Navi 1x by now. At least Navi 2x has RT, and reportedly is still being sold! I'm disappointed with this, because even Navi 2x can run very new stuff good. Especially RX 6750 XT and the RX 6800 series.

A secondary driver branch or legacy, whatever anyone would choose to call it, doesn't mean driver support is ending.

AusWolfAh, I get what you mean. But still, the question remains: why do you think this is a problem, or will be a problem in the future?

Isn't it obvious? It means that the kernel mode components will no longer be updated. This means that newer extensions plus any and all low-level functionality that interfaces directly with the hardware will no longer be added or developed. Bug fixes are as far as you'll get there. Optimizations for performance will be limited to what's feasible with that stack. Effectively, it's a second-class citizen at that point.

Dr. DroIsn't it obvious? It means that the kernel mode components will no longer be updated. This means that newer extensions plus any and all low-level functionality that interfaces directly with the hardware will no longer be added or developed. Bug fixes are as far as you'll get there. Optimizations for performance will be limited to what's feasible with that stack. Effectively, it's a second-class citizen at that point.

No new feature is coming for RDNA 1 and 2, that was fact way before the ominous announcement.
Bug fixes and new game support will be uninterrupted, AMD said that. Also, what bugs are we talking about a mere 5 years after the RDNA 2 launch? Have they not been fixed already?
Performance optimisations... what performance optimisations are you expecting on a 5 year-old architecture that haven't been implemented a long time ago?

Maintenance mode or not, RDNA 1 and 2 aren't gonna improve with new drivers. AMD has done everything they could to optimise them as much as possible. If they won't update their kernel components as often as they do for RDNA 3 and 4, so what? You're not losing anything - your RDNA 2 card would behave 100% identically even with those kernel updates as it does now. It's wasted effort.

TL,DR: You're talking like a new driver could turn around anyone's user experience on a 5 year-old architecture. Like there was still a lot of performance to be extracted by drivers, and a lot of bugs to be fixed... again... 5 years after release... Like AMD did absolutely nothing in these 5 years driver-wise, and only a 2026 version could make a 6700 XT work as intended. Like drivers were magic or something.

AusWolfNo new feature is coming for RDNA 1 and 2, that was fact way before the ominous announcement.
Bug fixes and new game support will be uninterrupted, AMD said that. Also, what bugs are we talking about a mere 5 years after the RDNA 2 launch? Have they not been fixed already?
Performance optimisations... what performance optimisations are you expecting on a 5 year-old architecture that haven't been implemented a long time ago?

Maintenance mode or not, RDNA 1 and 2 aren't gonna improve with new drivers. AMD has done everything they could to optimise them as much as possible. If they won't update their kernel components as often as they do for RDNA 3 and 4, so what? You're not losing anything - your RDNA 2 card would behave 100% identically even with those kernel updates as it does now. It's wasted effort.

TL,DR: You're talking like a new driver could turn around anyone's user experience on a 5 year-old architecture. Like there was still a lot of performance to be extracted by drivers, and a lot of bugs to be fixed... again... 5 years after release... Like AMD did absolutely nothing in these 5 years driver-wise, and only a 2026 version could make a 6700 XT work as intended. Like drivers were magic or something.

Yeah dude tell that to the people who bought an RTX 2060 almost 7 years ago and are running DLSS 4 super resolution and ray reconstruction today

It's as the old saying goes you can take a horse to the river but you can't make it drink, you pretty much elected to ignore the implications and forgive them for it

Dr. DroYeah dude tell that to the people who bought an RTX 2060 almost 7 years ago and are running DLSS 4 super resolution and ray reconstruction today

No new driver would give RDNA 2 users FSR 4 if that's what you're implying. Just like 1080 Ti owners will never get any version of DLSS, regardless of having a new driver installed or not.

You're either deflecting the topic into another one, or confusing the two. FSR 4 support has nothing to do with the "maintenance mode" announcement.

Dr. DroIt's as the old saying goes you can take a horse to the river but you can't make it drink, you pretty much elected to ignore the implications and forgive them for it

There are no implications, and you're using that old saying in the wrong context.

Hecate91I wouldn't trust Intel to keep arc going, given that they've taken an investment from Nvidia.

A secondary driver branch or legacy, whatever anyone would choose to call it, doesn't mean driver support is ending.

I think the concerning thing regarding RDNA2 here is that it is being lumped in with RDNA1 on the exact same timeline despite having significant hardware differences. I don't buy that it is the lack of hardware that is the reason behind degrading the support because AMD has the same end date for both.

If this was a case where RDNA2 got an extra year or whatever before its period ends, that would make more sense as an engineering challenge. Currently, it just looks like a typical anti-consumer move and I will die on this hill about this point.

It is a real pity that Intel could not get its affairs in order, this market desperately needed a third vendor to provide more competition.

AusWolfNo new driver would give RDNA 2 users FSR 4 if that's what you're implying. Just like 1080 Ti owners will never get any version of DLSS, regardless of having a new driver installed or not.

You're either deflecting the topic into another one, or confusing the two. FSR 4 support has nothing to do with the "maintenance mode" announcement.


There are no implications, and you're using that old saying in the wrong context.

Right, this is why somehow users managed to get FSR4 working unofficially on RDNA 2, because it is impossible for AMD to do so. What are you saying, that AMD's driver division is more incompetent than some random nerd on the Internet? :wtf:

I choose to believe that it is AMD being greedy here, not some complicated technical reason why it is unfeasible. It goes beyond the suspension of disbelief to assume AMD is acting in complete good faith here. What I find confusing is why you can't seem to see that AMD might, just might not have YOUR or MY interests at heart. This is Capitalism 101 here.

TechBuyingHavocRight, this is why somehow users managed to get FSR4 working unofficially on RDNA 2, because it is impossible for AMD to do so. What are you saying, that AMD's driver division is more incompetent than some random nerd on the Internet? :wtf:

I choose to believe that it is AMD being greedy here, not some complicated technical reason why it is unfeasible. It goes beyond the suspension of disbelief to assume AMD is acting in complete good faith here. What I find confusing is why you can't seem to see that AMD might, just might not have YOUR or MY interests at heart. This is Capitalism 101 here.

For some reason, AMD decided not to make a version of FSR 4 that runs on RDNA 2. It's sad, I give you that. I just don't see how not relegating drivers into maintenance mode would change that.

AusWolfNo new driver would give RDNA 2 users FSR 4 if that's what you're implying. Just like 1080 Ti owners will never get any version of DLSS, regardless of having a new driver installed or not.

You're either deflecting the topic into another one, or confusing the two. FSR 4 support has nothing to do with the "maintenance mode" announcement.

AusWolfFor some reason, AMD decided not to make a version of FSR 4 that runs on RDNA 2. It's sad, I give you that. I just don't see how not relegating drivers into maintenance mode would change that.

:confused: What are you saying? The driver is what enables FSR support. How does a new driver NOT give FSR4 to RDNA2 if that is what AMD designs the driver to do? How does putting a driver into maintenance mode NOT close off the possibility of getting FSR4 support for RDNA2?

These two things are the one and the same, the change of status into maintenance mode means AMD is telling us that no FSR4 is coming for RDNA2, despite it being trivially easy to accomplish.

TechBuyingHavocsome AMD defenders are trying to pigeon-hole AMD complainers as Nvidia-only consumers to reduce the legitimacy of their statements.

It's a double doozey isn't it, it's already exceptionally poor reasoning to assert those who dare to have a not-positive opinion on this topic would never buy AMD anyway, and then layer on it fallaciously concluding that due to that, the opinion has no legitimacy at all. The sheer hubris of it.

TechBuyingHavoc:confused: What are you saying? The driver is what enables FSR support. How does a new driver NOT give FSR4 to RDNA2 if that is what AMD designs the driver to do?

Because AMD has not, and will not design the driver to do that.
How many driver versions have been released since FSR 4 was introduced? Did any one of them enable it on RDNA 2? No. Not a single one of them. So I don't get why you'd assume that any future non-maintenance mode driver would do that. AMD categorically said that FSR 4 is not coming for older architectures. You can think about it what you will, but non-maintenance mode drivers wouldn't change that.

TechBuyingHavocHow does putting a driver into maintenance mode NOT close off the possibility of getting FSR4 support for RDNA2?

How does not putting drivers into maintenance mode guarantee FSR 4 support for RDNA 2?

TechBuyingHavocThese two things are the one and the same, the change of status into maintenance mode means AMD is telling us that no FSR4 is coming for RDNA2, despite it being trivially easy to accomplish.

They told us that way before they put driver development into maintenance mode. Non-maintenance mode drivers would change nothing.

wolfIt's a double doozey isn't it, it's already exceptionally poor reasoning to assert those who dare to have a not-positive opinion on this topic would never buy AMD anyway, and then layer on it fallaciously concluding that due to that, the opinion has no legitimacy at all. The sheer hubris of it.

I'm not suggesting anyone to have a positive opinion about no FSR 4 support on RDNA 2.
What I'm saying is that maintenance mode, or non-maintenance mode drivers do not change the fact. These are two totally unrelated topics.

Redstone, 10th December - we'll see how RDNA2/3 fares

TheinsanegamerNI cant sum it up better than the article did.

I bought a 6800xt less than 4 years ago. I intentionally bought an AMD card because my intention was to run Linux as my main OS. I waited over a year through unavailability and finally got the chance to buy one. Months of using a vega 64 that had to be run at ever shrinking power targets to maintain stability.

And this is my reward.

Really puts a harsh buzz killer on buying into uDNA when it comes out. nVidia has done the same, but only to the 2000 series from 2018, not the 3000s.

I read a post or news here or somewhere else,

Saying that developing RDNA 1 2 3 4 for AMD has been taxing because, the style of development was like a complete reset of something like at least optimisations, instead of the idea of architectural improvements over time, it's reset new one new techniques, then new one new again 4 times in a row, Idk about Nvidia.

Amd is managing 4 different architectures and likely is developing The fifth one GDNA already which they said isn't gonna be reset new reset new every further generation of that! but actually improving over time I think, so to sum it up I guess AMD couldn't handle holding all these architectures in game day optimisation, and I think it makes sense considering Nvidia was the one stable drivers and AMD was the unstable drivers before Intel, as in their driver developement capabilities we will see How Everything Goes W1zzard TheinsanegamerN and everyone else! btarunr too!

Benedict Luashfu Chen!!!!!!!!
AMD was serious price to performance during or a bit after crypto at that time in Australia, rx 6600 I saw msi card $329. I remember clearly such a low price for high performance!

Benedict Luashfu Chen I Repeat!!!!!

Let me just leave this here:

Some people don't understand branch splits.

dgianstefaniSome people don't understand branch splits.

Some people don't see and/or care about the difference.

AusWolfSome people don't see and/or care about the difference.

A RDNA 4 user who has a few years of support left (probably/maybe) doesn't see or care, or recognise the significance of their RDNA 2/3 brethren being placed on a low priority separate branch. To the point where FSR Redstone gave them literally nothing. For products still being sold.

dgianstefaniA RDNA 4 user who has a few years of support left (probably/maybe) doesn't see or care, or recognise the significance of their RDNA 2/3 brethren being placed on a low priority separate branch. To the point where FSR Redstone gave them literally nothing. For products still being sold.

FSR Redstone was never promised for RDNA 2 and 3. In fact, it wasn't even news during their release. How could anyone have expected it?

Similarly, are you 100% certain that whatever Nvidia comes up with in the future will run on your Blackwell card?
Did you ever buy a Pascal card thinking that you'll run RT and DLSS on it? I didn't, but I still bought several of them. I still have a 1050 Ti and a 1030 working just fine.

You're talking like you had a crystal ball that tells you about the future support of your card. Truth is, you don't. Nobody does.

Let's not put on the old record, shall we? ;)

I only linked the above to show that despite FSR Redstone not being available for RDNA 2 and 3, the "end of software support" fear mongering has no basis in reality.

AusWolfSimilarly, are you 100% certain that whatever Nvidia comes up with in the future will run on your Blackwell card?
Did you ever buy a Pascal card thinking that you'll run RT and DLSS on it? I didn't, but I still bought several of them. I still have a 1050 Ti and a 1030 working just fine.

You're talking like you had a crystal ball that tells you about the future support of your card. Truth is, you don't. Nobody does.

Stop trying to shift the focus away from the fact that RDNA 2/3 are no longer on the priority branch for drivers, that branch includes RDNA 4, and nothing else.

AusWolfFSR Redstone was never promised for RDNA 2 and 3. In fact, it wasn't even news during their release. How could anyone have expected it?

Because AMD's competition runs their latest upscaler on Turing, a 2018 architecture. Some parity is expected believe it or not.

dgianstefaniNVIDIA is irrelevant to this discussion, despite their much better support track record.

As long as you keep comparing AMD to them, they're unfortunately not irrelevant.

dgianstefaniStop trying to shift the focus away from the fact that RDNA 2/3 are no longer on the priority branch for drivers, that branch includes RDNA 4, and nothing else.

Like I said, who cares? I merely showed that new drivers are still being made for them despite the panic being spread ever since that blunderous announcement.

dgianstefaniBecause AMD's competition runs their latest upscaler on Turing, a 2018 architecture. Some parity is expected believe it or not.

But not multi-frame generation. Oh I know, that's not important now, is it? :rolleyes:

AusWolfAs long as you keep comparing AMD to them, they're unfortunately not irrelevant.

AMD is the comparison, NVIDIA is the standard, AMD have tiny DGPU marketshare, and keep digging the hole with stuff like this. One generation old GPUs having limited support is just the latest example of a underdog second place mentality, and people who defend this behaviour out of brand loyalty don't make things better.

AusWolfLike I said, who cares? I merely showed that new drivers are still being made for them despite the panic being spread ever since that blunderous announcement.

11 pages of people care. Plus many others.

dgianstefaniAMD is the comparison, they have tiny marketshare, and keep digging the hole with stuff like this.

Why do you keep shifting the topic? I only posted a driver update/support page. :kookoo:

Why do you have to make every tiny thing into an overblown argument? :(

dgianstefani11 pages of people care. Plus many others.

That's why I posted the above.

AusWolfWhy do you keep shifting the topic? I only posted a driver update/support page. :kookoo:

Why do you have to make every tiny thing into an overblown argument? :(

If you intend to mislead and muddy waters, implying "current" drivers working means there's nothing to worry about, when in fact the cards are on a separate driver branch within that package and already aren't receiving feature updates, someone will always be there to clarify things for those that might otherwise be misled.

Like it or not, that's on topic.