You have to prioritize among support, accessibility, features, image quality etc. All those matter and you can't say DLSS has better image quality thus it will be the main focus for devs. DLSS is very good but can the scale lean towards DLSS just because it has a bit better image quality and few frames more in some cases but it cannot be used broadly on wide range of devices? I will skip even implementation but proprietary software is not for everyone.
Considering they are so similar, and that Nvidia is themselves making it easier to implement all similarly working reconstruction (FSR, XeSS and DLSS) all at once (streamline), I don't think this will be the barrier people assume (and often hope) that it is. More below.
If AMD gets more market share (everything suggests it will happening) FSR will be the better option as a main focus.
Maybe, and again I am absolutely not saying DLSS will be, but we are discounting any other options from entering the market, better universal options that aren't vendor locked or even vendor
labelled. And even if it was, it still is extremely easy at that point to also add DLSS for the PC market.
What is also very important which you have missed to consider. The more products can use a feature the more support for it and faster development. You may say DLSS has better support and more games support and it is true. what about acceleration with which these are being implemented and used? How about consoles? These are gaming devices and DLSS and FSR can improve things here as well right? Which one of these two will be the main focus? (Not even main but only one focus). How is porting games from PC to Consoles and vice versa coming on nowadays? What is the market of consoles vs PC? Now tell me which aspect is more important here, accessibility or image quality? How great image quality of a feature must have been so that accessibility would not matter?
People do like to talk a lot about market shares and such, but there are also multiple ways to consider that, as well as the ease of cross-implementation as mentioned above. Sure, AMD has consoles fairly cornered, and are quite possibly on the upswing as far as Desktop GFX card market share goes, but Nvidia still have a massive PC market share, a presently continuous circa 80% sales share of modern cards too, it's been about 4 years now that 4/5ths of PC GFX cards sold have RTX capability.
Naturally, you look at something like steam + all consoles in existence and it's easy to say Nvidia don't have majority share there so why imp? it's still the vast majority of the current market for AAA PC gaming and multiple millions of people who would want DLSS too, or at least benefit from it's inclusion relative to the ease to implement, and given the majority of the work needed for either is the same/similar/aided by streamline, I would wager on both for a while. Accessibility counts for a tonne, more than IQ even sure, but when another, superior option exists and the barrier to entry is extremely low, assuming one or the other gets implemented, the minimal amount of extra work does still make sense for game developers. We're not talking about months of work here, when one is in it's days tops. Perhaps we are actually mildly agreed on this point.
No one says DLSS wont be there.
Oh I could certainly show you people banking on it's death inside 1-2 years, some spurred on by a article title used on this very site.
my 2c are, I expect FSR to continue to gain ground on DLSS, both IQ wise, performance penalty wise, and adoption wise, in the short to mid term, and yet-unseen options to evolve or enter the market as lasting universal techniques that become what you could consider de-facto reconstruction. This could be FSR, but I don't think it will be FSR
2.0 - perhaps a later, better revision, I don't have a crystal ball, as it seems so many people do given the certainty they talk with. Such is the way innovation works, your average consumer cannot innovate, they just demand more, better faster - but we don't know what we don't know. I absolutely do not see dev's thinking, well FSR is kinda close, 'good enough' even, and despite doing the majority of the work, lets skip adding DLSS too. This seems to be where I disagree most with people - whether they argue from the mind of from the heart. Nvidia creating streamline to make it easy to add them all is very telling, they will actively develop and provide the tools to keep DLSS in the game for now.
Maybe we agree a little, but disagree a lot too, and I'm not sure what's left to be said to each other on this topic.
