NVIDIA is full of shit
Technology
118
Beiträge
53
Kommentatoren
188
Aufrufe
-
But nvidia got dragged across the coals for using frame-gen in their performance benchmarks too. Did you miss that?
Also ATI wasn't owned by AMD then.....AMD aquired ATI in 2006. Your link is from 2001.
Also no one should be listening to official GPU manufacturer benchmark results. No one. Review companies do their own benchmarking, and you do know that you can turn off DLSS and DLSS Frame-Gen, don't you? I haven't seen any reviewers only compare DLSS+Frame-Gen on an nvidia card to native-with-no-frame-gen on AMD cards. You must have, so can you link to any?
@FreedomAdvocate so you didn't read the heise link which showed you that pre release tests had strict rules on how to test including framegen settings ...
-
Second cuda is not hardware dependend
That's essentially an emulation layer. Nvidia make DLSS specifically for their GPUs, which have CUDA cores on them. It's the reason why DLSS doesn't work on their pre-CUDA core hardware.
Could they make DLSS work on AMDs hardware? Sure, they could - but it would not be DLSS as we know it, and again - why would they? They are allowed to make stuff exclusively for their hardware.
@FreedomAdvocate zuda is an reimplementation of an api not a emulation.
-
I 100% know what DLSS is, though by the sounds of it you don't. It is "AI" as much as any other thing is "AI". It uses models to "learn" what it needs to reconstruct and how to reconstruct it.
No, you don't. https://en.m.wikipedia.org/wiki/Deep_Learning_Super_Sampling
This is blatantly and monumentally wrong lol. You think it's literally rendering a dozen frames and then just picking the best one to show you out of them? Wow. Just wow lol.
Literally in the docs: https://raw.githubusercontent.com/NVIDIA/DLSS/main/doc/DLSS_Programming_Guide_Release.pdf
What it does is allow you to run a game at higher settings than you could usually at a given framerate, with little to no loss of image quality. Where you could previously only run a game at 20fps at 1080p Ultra settings, you can now run it at 30fps at "1080p" Ultra, whereas to hit 30fps otherwise you might have to drop everything to Low settings.
No it doesn't. It allows you to run a game at a higher resolution for no reason at all, instead of dropping to a lower resolution that your card can handle natively. That's it.
Keep claiming otherwise, and you're just literally denying reality and the Nvidia link to the docs right in front of you.
Linking to an 81 page document isn’t helpful. What specifically in there are you referring to?
No it doesn’t. It allows you to run a game at a higher resolution for no reason at all
Other than the reasons like I said - running it at higher settings while maintaining a playable framerate. The point is you don’t have to lower settings as much with DLSS.
You fundamentally don’t understand what it is and what it allows you to do.
-
@FreedomAdvocate so you didn't read the heise link which showed you that pre release tests had strict rules on how to test including framegen settings ...
Nvidia can say what they want, but reviewers didn’t follow those.
Sounds like you need to find better GPU review sites.
-
@FreedomAdvocate zuda is an reimplementation of an api not a emulation.
I said it’s essentially emulation, which it is. Its like WINE, which is also essentially emulation but isn’t emulation.
-
Linking to an 81 page document isn’t helpful. What specifically in there are you referring to?
No it doesn’t. It allows you to run a game at a higher resolution for no reason at all
Other than the reasons like I said - running it at higher settings while maintaining a playable framerate. The point is you don’t have to lower settings as much with DLSS.
You fundamentally don’t understand what it is and what it allows you to do.
You didn't read up on it huh?
Hilarious.
-
I said it’s essentially emulation, which it is. Its like WINE, which is also essentially emulation but isn’t emulation.
@FreedomAdvocate there is a reason why WINE = Wine is not (a) Emulator is used. So don't call a api reimplementation a emulation specially since other api reimplementation have shown to be better than the original implementation from the hardware provider ( example dxvk on amd > the original amd dx implementation ) . But this gets us far from the original topics , my point was if nvidia wanted to have real competition they would have included all those new fance features into official api's like for example DX or Vulkan or any other.
They didn't ... and while not directly against the consumer it is against the consumer end.
So i have brought up another point why i call nvidia anti consumer ... neither you like it or not. -
This post did not contain any content.
You don't need NVENC, the AMD and Intel versions are very good. If you care about maximum quality you would software encode for the best compression
-
Since when did gfx cards need to cost more than a used car?
We are being scammed by nvidia. They are selling stuff that 20 years ago, the equivalent would have been some massive research prototype. And there would be, like, 2 of them in an nvidia bunker somewhere powering deep thought whilst it calculated the meaning of life, the universe, and everything.
3k for a gfx card. Man my whole pc cost 500 quid and it runs all my games and pcvr just fine.
Could it run better? Sure
Does it need to? Not for 3 grand...
Fuck me!.....
When advanced nodes stopped giving you Moore transistors per $
-
I plan on getting at least a 4060 and I'm sitting on that for years. I'm on a 2060 right now.
My 2060 alone can run at least 85% of all games in my entire libraries across platforms. But I want at least 95% or 100%
Why not just get a 9060 xt and get that to 99%?(everything but ray tracing black myth wukong)
-
Then why does Nvidia have so much more money?
Because of vendor lock in
-
Like I said...you don't know what DLSS is, or how it works. It's not using "AI", that's just marketing bullshit. Apparently it works on some people
You can find tons of info on this (why I told you to search it up), but it uses rendering tables, inference sorting, and pattern recognition to quickly render scenes with other tricks that video formats have used for ages to render images at a higher resolution cheaply from the point of view of the GPU. You render a scene a dozen times once, then it regurgitates those renders from memory again if they are shown before ejected from cache on the card. It doesn't upsample, it does intelligently render anything new, and there is no additive anything. It seems you think it's magic, but it's just fast sorting memory tricks.
Why you think it makes games better is subjective, but it solely works to run games with the same details at a higher resolution. It doesn't improve rendered scenes whatsoever. It's literally the same thing as lowering your resolution and increasing texture compression (same affect on cached rendered scenes), since you bring it up. The effect on the user being a higher FPS at a higher resolution which you could achieve by just lowering your resolution. It absolutely does not make a game playable while otherwise unplayable by adding details and texture definition, as you seem to be claiming.
Go read up.
It actually doesn't do what you said anymore, not since an update in like 2020 or something. It's an AI model now
-
He’s talking about DLSS upscaling - not DLSS Frame Generation - which doesn’t add latency.
It does add latency, you need 1-2ms to upscale the frame. However, if you are using a lower render resolution (instead of going up in resolution while rendering internally the same) then the latency will be lower because you have a higher frame rate
-
After being AMD for years recently went back to nvidia for one reason. nvenc works way better for encoding livestreams and videos than amd
Fixed in 9000 series
-
Why not just get a 9060 xt and get that to 99%?(everything but ray tracing black myth wukong)
Why not - no?
I don't even care for the monkey game.
-
It does add latency, you need 1-2ms to upscale the frame. However, if you are using a lower render resolution (instead of going up in resolution while rendering internally the same) then the latency will be lower because you have a higher frame rate
Yeah, so it doesn’t add latency. It takes like 1-2ms iirc in the pipeline, which like you said is less than/the same/negligibly more than it would take to render at the native resolution.
-
@FreedomAdvocate there is a reason why WINE = Wine is not (a) Emulator is used. So don't call a api reimplementation a emulation specially since other api reimplementation have shown to be better than the original implementation from the hardware provider ( example dxvk on amd > the original amd dx implementation ) . But this gets us far from the original topics , my point was if nvidia wanted to have real competition they would have included all those new fance features into official api's like for example DX or Vulkan or any other.
They didn't ... and while not directly against the consumer it is against the consumer end.
So i have brought up another point why i call nvidia anti consumer ... neither you like it or not.I’m not sure if English isn’t your first language, or if you’re just being wilfully obtuse, but I didn’t call it emulation. I said it is essentially emulation, like WINE. I know WINE isn’t emulation, which is why I said it is “essentially” emulation because it’s doing the same thing - converting calls from one set of APIs to work on other hardware/architecture. It’s not emulation, but it’s essentially the same thing.
Why would Nvidia want competition? AMD don’t want competition either, but they made FSR work on everything because they were so far behind Nvidia (and because it was all done in software, requiring no special hardware) that they have to give it away to try and catch up.
Companies making proprietary tech is not anti-consumer - unless of course you think that everything other than making everything free and open source is “anti-consumer”, which I am thinking you might?
-
You didn't read up on it huh?
Hilarious.
I know what DLSS is, I want to know exactly what you think in there says what you think it does.
So are you going to actually substantiate your claims with evidence/proof or no?
-
It actually doesn't do what you said anymore, not since an update in like 2020 or something. It's an AI model now
The docs I linked to LITERALLY explain what it is, how it works, and the mechanics behind it. Not until DLSS 4 released earlier this year do the docs mention any kind of way to employ the use of of models to anything at all.
-
Yeah, so it doesn’t add latency. It takes like 1-2ms iirc in the pipeline, which like you said is less than/the same/negligibly more than it would take to render at the native resolution.
Which also means it's not possible to use it to go to 1000 fps
