Srry, but you replay on my text then keep saying things that I or in fact no one in here ever said at least not that I am aware....So after all that was only yours internal thoughts and just some kind of theorycrafting?ahh now I get it....this also probably clear some of yours previous responds to my posting....
The reason I keep mentioning driver optimizations is that this is the
actual, tangible, real-world gain that users frequently get from driver updates. It is, in the grand scheme of things, a realistic expectation. In this case though, that's not relevant, as the gains will not be large enough. That's why it keeps getting mentioned. But you're right, you're not arguing that, so I'll stop as well. At least we agree on that point.
As for the rest ... uh, what are you talking about? Theorycrafting? I'm pointing out that the amount of bugs found on architectures that old are extremely small, and that any new major bugs are likely to come from entirely new game engines with new features that aren't supported by the architecture. Who knows how UE5 will work on these? But also, who cares how UE5 will work on cards that aren't likely to be able to run UE5 games at any kind of playable performance? And, as I've said before, it's relatively common to see hotfix drivers for major issues even with unsupported hardware. So if something breaks fundamentally, it's still likely to get fixed (or there might be an advisory to downgrade to a previous driver or similar). (It's also relatively unlikely that a new game engine will fundamentally break compatibility with a decade's worth of consoles and PCs unless it's implementing mandatory features that these architectures lack, such as RTRT.)
I mean, I asked you a very specific question: how many bugs related to the now-unsupported GPUs can you find in the past year? Are they game or feature specific, or are they more significant? And have they been fixed? If a) the number is low, b) the issues found are highly specific, and c) the issues haven't been fixed by now, then this announcement effectively changes nothing. You're the one insisting this is a real problem, thus the onus is on you to provide evidence to support this.
From a quick skim of the changelogs for previous drivers listed on AMD's site (going back to september), searching for "GCN", here's what I found:
- A system hang or crash may be experienced when upgrading Radeon Software while an Oculus™ VR headset is connected to your system on Radeon GCN graphics products
- Reported in 21.2.2, fixed in 21.3.1 (two versions later)
- Project CARS 3™ may experience mirror like corruption when using VR during game menus on GCN based Radeon graphics products.
- Listed in 20.9.1, can't find a mention of it as fixed, but it disappeared from "Known issues" after this. Likely fixed by a game update (or maybe an Oculus driver update)?
And that's it for mentions of GCN, at least. There are no mentions of "R9" or "Fury" outside of the compatibility lists. I didn't search of anything older. As such, it seems that a grand total
two issues relating to these cards have been found and addressed in the past 9 monts, one of which seemingly by a game update rather than a driver update, and both of them are highly niche and specialized. And given VR performance on the now-unsupported GPUs, it's quite unlikely that these were the GPUs on which the issue was found.
What I'm trying to get across here: the scenario that you are scared of is
extremely unlikely to come to pass. If no issues were found in the past 9 months - a period in which hundreds of games have been released! - why would they appear in the next 9? Or the 9 after that? And at that point, would it at all be worth discussing, given the actual performance of these GPUs?
after all....
