Edit: This video amongst other DF videos just proves once again, how far behind AMD is with RT performance. Heck, at times Intel's Arc is outperforming AMD's RT performance.
Last edited:
Digital Foundry Article Technical Discussion [2023]
- Thread starter BRiT
- Start date
- Status
Edit: This video amongst other DF videos just proves once again, how far behind AMD is with RT performance. Heck, at times Intel's Arc is outperforming AMD's RT performance.
You'd typically expect an x60 to be significantly faster than the x70 from the prior generation. Now, it not only is much slower than the previous x70, but it also has less VRAM than the previous x60 lol.
Dont understand the comparision to the PS5. Here in Europe the disc version costs 220€ more. The Xbox Series S would have been better because the price difference is less.
The minimum GPU support is 1070 TI on nVidia's side. There are only three Pascal GPUs official supported and they make less than 2,2% on Steam. On the other hand there are ~37% RTX GPUs faster than the 1070 TI. So nearly 80 million PC gamers will have a worse PC experience.
Last edited:
D
Deleted member 11852
Guest
Either Xbox would have been a more relevant comparison point because the consoles also use a variation of the DirectX API. I may be imagining it, but it feels like in a few videos that PS5 has become the default console comparison console ¯\_(ツ)_/¯
Because PS5 is what most console gamers will be playing on.
Flappy Pannus
Veteran
You need more than a GPU to actually play games. Nvidia is the one arguing that an equivalent GPU should cost the same as an entire console, which is ridiculous. 2 1/2 years into a consoles lifespan, a $300 GPU should certainly be comparable to a $400 console.
Last edited:
davis.anthony
Veteran
Most people already have a system though and just need a GPU upgrade.
Well, I'm one of the people with an RTX GPU, and I love DLSS, but I think I'll wait until the game comes out before I start crying about how bad my experience is. And then I'll wait until a DLSS mod comes out before I compare what we got to what could have been.
Good video. Seems like a really great PC version.
And great to see all the upscaling methods supported
davis.anthony
Veteran
Is it the first racing game with RTGI?
nice video. My favourite thing about Ray Tracing is that it depends on the GPU for it to work, not the device you connect the GPU to.
As of currently, I am playing on my 1080p 32" TV from 2013, which has great image quality -allegedly 12bits, though I dont believe that claim, but it looks neat-. I play on it 'cos I love to play in silence, and from 1440p and above that, the GPU works more and it can get noisy. The TV doesn't have HDR -it wasn't a thing 10 years ago- but I love that tradeoff as long as I can play in silence and the GPU rests in its laurels.
So I bit the bullet and finally ordered a 5800X3D to drop in my AM4 motherboard. Spiced it up with a Thermalright Peerless Assassin 120 SE to replace the AMD stock cooler too so I'm hoping for a much quieter system under load.
Can't wait to try out the various games that my 3700X was struggling to hit 60fps on!
gamervivek
Regular
It's quite apropos that Richard talked of this concept of hotspots in the 3700X review video, ( at 24m15s )
imo these hotspots are impossible to be seen on normal reviews, because they'd just a have runthrough of the place. Meanwhile if you're playing a single-player game and have to revisit this kind of hotspot again and again, you'd tear your hair out. Swan's Pond and adjacent areas in Fallout 4 were notorious for this and I saw a 50% increase going to 9900K from 3600.
Then there are some specific optimization issues for Ryzens, my 12700K was almost double of 5800X in some GTAV areas which was curiously due to water settings, despite water not being visible on the screen.
imo these hotspots are impossible to be seen on normal reviews, because they'd just a have runthrough of the place. Meanwhile if you're playing a single-player game and have to revisit this kind of hotspot again and again, you'd tear your hair out. Swan's Pond and adjacent areas in Fallout 4 were notorious for this and I saw a 50% increase going to 9900K from 3600.
Then there are some specific optimization issues for Ryzens, my 12700K was almost double of 5800X in some GTAV areas which was curiously due to water settings, despite water not being visible on the screen.
I've been thinking about doing the same. It's tough because there are games that really perform amazingly well with the bigger cache and you don't have to worry about memory tuning so much. Then you get those games that are massively single-thread bound and the clock of the 5800x3d means it's more limited than other cpus. Overall it's a massive win in the majority of cases, and I already have an AM4 board so it's my cheapest path. But I'm also not sure I'm happy enough with the games that are coming out to really care. I think Alan Wake 2 will probably be a determining factor for me.
I bought my AM4 board right around when zen 2000 series were coming out, but I got a 1700x on sale for dirt cheap. The gaming performance was ... not great. Then when all of the 3000 series stuff came out, a ton of reviews talked about how they'd really solved the gaming problem, so I grabbed a 3600x on sale at some point while I was heavy into Apex Legends. When I got it at first I was really happy, but noticed there were spots on the Apex maps where my fps tanked, and my friends with older Intel CPUs weren't having any issues. The performance of the 3600 just seemed kind of all over the place. It was generally good, but when it was bad it was very bad. Eventually I got a 5600x on sale, and I'd say it's considerably more stable than a 5600, but I think it's still true that if you run an intel cpu for gaming you just won't suffer the same performance issues in demanding areas. The 5800x3D is probably the first really incredibly good gaming CPU that AMD put out. Unfortunately I don't think the 7800x3d, or whatever it is, has the same great value. If I don't abandon AAA games, I might go intel for my next platform if current trends continue.
tested this some time ago when I installed a mod.
Resident Evil 4 Remake FSR2 Quality vs XeSS Quality - Imgsli difference here is more pronounced.
Vegetation density and texture definition are the main differences I see. This is what you get when you prohibit other techs from being used in games.
Resident Evil 4 Remake FSR2 Quality vs XeSS Quality - Imgsli difference here is more pronounced.
Vegetation density and texture definition are the main differences I see. This is what you get when you prohibit other techs from being used in games.
Flappy Pannus
Veteran
Interesting, but I think Oliver could have been a little clearer at the start in emphasizing that this toolkit is really meant to be used by developers as a bridge to get native, or pseudo-native games on the Mac. He mentions its similar to Proton, it kind of is in how it operates, but the goals are decidedly different. Proton is meant for Linux/Steam Deck users to play Windows games natively OOTB, this is not that...yet.
Also not mentioned in the video title, but at the end there's a quick look at the native No Man's Sky port on Mac, particularly how its implementation of MetalFX Temporal Upsampling fares against DLSS and FSR when reconstructing to 1080p (spoiler: Very close to DLSS in quality, FSR2 far behind).
Last edited:
- Status
Similar threads
