Nvidia is a more premium product. Better and longer driver support and featureset. AMD just abandoned RDNA 2 and RDNA 3 with old crappy upscaler while an entry level nvidia gpu from 2018 has the best upscaler in the market. Clearly it's a more premium experience.
It's not brand name. It's support. I'm sure you made a similar comment back in 2018 when people preferred a 2070 over the 5700xt. 6 years later the people that bought the 2070 are laughing all the way to the bank, having access to the latest and greatest upscaler for the past 8 years while the people with the 5700xt have been abandoned, were forced to upgrade (some games didn't even launch on the 5700xt) and the amd card they upgraded to has been also abandoned. It's not brand that's selling, it's consistently offering mroe longevity and support for their products .
Premium for what? Too expensive for actual gameplay, premium for upscaling and fake frames?
Premium on offering support for much longer. Again, a 2018 entry level nvidia card had the best upscaling on the market for 8 consecutive years, much better than a 2022 flagship 1k$ card from AMD has. That's premium.
Is that some sort of a Strawman argument? Jokes aside, my Asus 1080TI died and the refused to warranty it. Nothing about my 7800gt, 8800gt sli, gtx280, it 580 sli, 980 or 1080ti was a luxury experience. I run a 6700xt in my main rig because my 1080 died during covid aand best I could replace it with was a 1050ti for awhile. I have a 3070ti in my AI rig and a 2060 in the wife's PC.
Is driver and feature support not an indication of premium? What would qualify as premium to you then? Cause under my definition the 2070 is more premium than the 5700xt on every account.
So does this "premium upscaler" include multi frame gen for the RTX 20 series?
AMD didn't abandon RDNA2 or 3, they're getting FSR4.1 as well.
www.techspot.com
Unless you want to look at Linux support, nVidia drivers have been sh*t the last year and that's mostly just cuda and AI
Don't worry about that. Jensen has been so ****ing high on stimulants in his latest interviews that I don't think he can pilot the ship much longer
You are dodging the question repeatedly. An entry level 2018 nvidia gpu gets a lot better support than a 2022 super high end amd flagship. Why don't you address that before you go off on a tangent?
Getting FSR 4 for my RDNA 2 in end of 2027 is useless. Why not get it, you know, last year alongside RDNA 4, when those gpus were still relevant?AMD didn't abandon RDNA2 or 3, they're getting FSR4.1 as well.
FSR 4 upscaling is finally coming to older Radeon GPUs, and the image quality gains are real
AMD's gaming and graphics VP, Jack Huynh, recently confirmed that Radeon RX 7000 graphics cards will begin supporting FSR 4.1 this summer, with support for RX 6000...
Though I wouldn't consider not having enough VRAM to enjoy the latest games, while having to upscale from 720p to make games playable a premium experience.
Quarterly results are plenty relevant when they consider gaming so unimportant they throw it into "edge computing". I couldn't imagine dropping $5k onto a 5090 to get told my purchase matters so little they won't even admit what gaming revenue actually is.
What question? Jensen has straight up said their drivers are mostly written by AI now. Their drivers a year ago are more stable than they are now. As far as support goes, Xbox and playstation both run on AMD and already have the optimizations written into them. Console gaming is bigger than PC gaming so AMD actually has some really great gaming support. Considering that AMDs stuff is mostly open source, you can literally copy and paste FSR4 support into basically any game that doesn't have it. Unless some developer wants to be a straight up **** with DRM and all that crap, but I don't pay for those games anyway because I don't give my money to *****
The question of why does an entry level 2018 nvidia card gets a better upscaler than a 2022 1k$ amd flagship. If that's non an indication of premium then what is?
Oh, that's easy! It doesn't. The 2018 GPU can't take advantage of all of DLSS's features and is basically stuck using DLSS 2. Don't worry, though, it can use FSR as that is a mostly open standard
Why are you lying? What possible benefit do you get out of this? First of all - DLSS 2 is better than FSR, so FSR is irrelevant anyways. Second of all, the lowly RTX 2060 can actually use the latest upscaler, DLSS 4.5 transformer model and RR. On the other hand, the newer and much more expensive RDNA 2 and RDNA 3 lineup cannot use the latest FSR. What possible reason is there for that other than amd scamming people?
The best parts of DLSS4.5 are hardware locked to the 50 series. The 2060 is using it in name only, hate to break it to you. it doesnt have the instruction set built onto the chip to run them. same with AMD. Nvidia is literally doing what you are accusing AMD of doing. They're actually bith doing it only AMD doesn't have hoards of fanboys trying to justify overpaying for hardware to get the same experience
The best part of DLSS 4.5 (the only part actually of DLSS 4.5) is the new transformer model and it runs on the 2060. You are making stuff up. MFG which you are obviously talking about was a DLSS 4 feature.
At a significant performance impact because the 20 and 30 series don't physically have the instructions on them to run it properly
So we went from not being supported to now being supported. Man, what the hell...
I wouldn't call that supported because I can hack fsr4 onto any GPU I want. I can emulate ARM on an x86 CPU, that doesn't mean it's "supported".So we went from not being supported to now being supported. Man, what the hell...
There is no performance penalty because you don't have to run the same quality. DLSS T at performance looks better than CNN at quality and therefore you get the advantages of both higher performance and higher image quality.
But that's all irrelevant frankly. None of that explains why low end 2018 nvidia cards get the feature while latest and greatest flagships from AMD do not. Let's totally ignore the elephant in the room my man
It isn't useless, and the assumption can be drawn that AMD and Sony are working together on FSR, RDNA2 coming later because consoles also use RDNA2.
I'm just curious, how much is Nvidia paying you?
DLSS 4.5 includes MFG and dynamic MFG, but Nvidia only allows it on the 50 series, not even allowing 40 series users to try it.
You having to hack fsr 4 means that amd doesn't officially support it cause they don't care. Nvidia does support it on their 2018 entry level gpus. Sotp ignoring the elephant in the room.
So consoles apparently also use rdna3, cause rdna 3 isn't getting it either. Great excuse.
I'm saying that I can, not that I should. I tried using the example of emulating ARM on X86 as an example. In many cases, trying to run DLSS4.5 on a 20 or 30 series card leads to a decrease in performance, not an increase. Yeah, it runs and if that's what you want to call support because you overpaid for Jensen's Kool-aid, go ahead and keep drinking.
AMD has admitted to what I said above about Nvidia. they can make it run, but since RDNA 3 doesn't have the instructions in it to run FSR4 properly, it can lead to a decrease in performance.So consoles apparently also use rdna3, cause rdna 3 isn't getting it either. Great excuse.
AMD drivers have been significantly worse, major stuttering in new games (forza horizon), blue screens, the cables catching fire (sapphire Asrock etc.) it's been a disaster of a release.
The best part of dlss 4.5 is the upscaler which works on entry level 2018 gpus. But it doesn't work on 2023 amd flagship. Let's keep ignoring the elephant in the room
