There is a wipeout VR for ps4. But yea i’d want a ’next-gen’ wipeout too.
Digital Foundry Article Technical Discussion [2022]
- Thread starter BRiT
- Start date
- Status
There is a wipeout VR for ps4. But yea i’d want a ’next-gen’ wipeout too.
D
Deleted member 2197
Guest
Wuuuutttt! switch! Goooooo! Love the flex here.
Reinforces the idea that GPU feature set plays a large role in game support more so than just raw power.
A great game, and surprisingly good performance and options given the size of the game and the first release of this developer. Definitely inspiring for smaller dev teams.
Around 10gb vram.
The info from their written article @ https://www.eurogamer.net/digitalfo...-memory-boost-and-the-massively-fast-rtx-4090
good discussion here. Nice to see the return of dee eff triple after the summer break. Alex is right about the version differences, turn up something else that isn't resolution. Especially as you get into that higher dynamic resolution, it's so hard to tell anyway without analysis.
betternatethanlever
Newcomer
Can current consoles ever get something like DLSS? Or rather, can FSR ever improve to come even remotely close to DLSS 2.0 original in terms of IQ? That would be a complete game changer.
Still wonder why series s still goes below 1080p in so quite some titles. It is more a PS4 pro so why does it dip so low. Settings are also almost always turned back. It more or less seems like lack of optimization so far. Don't think that a bit more memory will make such a big difference but it is always good to have more. Still don't understand why the os needs so much. After all xb360 is used just a few megabytes ...
Developers really must learn again how to reduce the footprint of software ... But we life in a world where "hallo world" needs 200+MB of main memory with a 2gb download size
I get your point, but to be fair ...
Lots of things done today that wasn't done in X360 days like running the entire Hypervisor for the host in order to run the Games and Apps containers as well as video encoding for recording gameplay or streaming. I'm surprised that they were able to cut 512 Meg off for Series X and and entire 1024 Meg for Series S in comparison to the Xbox One. They Series OS using 1 GB and 1.5 GB less than PlayStation 5 OS uses and that platform doesn't have any Hypervisor going on.
FSR 2.0 is arguably already pretty close, and it should improve.
Intels XeSS - not released yet and uses machine learning like DLSS 2.0.
It can run on any dp4a supported gpu which XS is.
PS5 may be able to run it, but would take some amount work to port it and wouldn't be as performant, which doesn't mean it wouldn't be worth it, we don't know.
Personally not expecting it.
Consoles have had TAAU for a while now, so not as big a game changer as you probably think.
But all these TAAU 2.0 (FSR 2.0, XeSS, TSR) will democratise good TAAU so will definitely be a big plus!
I think the potential performance advantage of DLSS over current console upscaling methods is if it can offer a visual presentation that is just as good or better than the console presentation while using a lower internal resolution.
One could argue though that at a fixed performance target on consoles that is held solid by DRS then on PC that would just translate into slightly better image quality.
Although the PC gamer could also elect for higher visual settings with similar visual quality and performance (assuming a console level GPU).
One could argue though that at a fixed performance target on consoles that is held solid by DRS then on PC that would just translate into slightly better image quality.
Although the PC gamer could also elect for higher visual settings with similar visual quality and performance (assuming a console level GPU).
Yep, I don't disagree.
I was just talking IQ, at the moment you can probably get more performance for the same IQ from DLSS.
Simply by running DLSS at a lower resolution.
At 1440p they are all close in IQ to looking like 4K now, DLSS can go lower though.
If what's shown here in this video is representative (posted by @snc in the other tech thread), then it looks like there are still types of detail where DLSS rules supreme over other reconstruction techs. Insomniac's technique appears to come in last, but I guess I should know more about what "balanced" means for each of them before I say that with any high degree of confidence.
DLSS continues to offer the most optimistic view of future reconstruction technology, IMO.
snc
Veteran
Also apparently dsr2 is better and there was opinion consoles doesn't need much dsr2 as they already use similar reconstruction techniques
- Status
Similar threads
