Inuhanyou
Veteran
I always find it interesting how console compares directly to PC like this, that's amazing
Current Generation Games Analysis Technical Discussion [2022] [XBSX|S, PS5, PC]
- Thread starter BRiT
- Start date
How would the ps5 be vastly faster than a 5700xt in this title? PS5 is 28% faster than the non XT, the 5700xt should be around 20% faster than the vanilla rx5700 which would be pretty close to ps5, which is exactly according to spec (9.7 vs 10.2TF).
Scaling almost exactly according to specifications, ps5 between 2070 to 2070S, xsx close to 2080 (even though he didnt test this with the new patch?).
Untill you start ray tracing ofcourse things would look different or go dlss vs the reconstruction used on consoles.
Game is supposedly gpu limited so we got a nice benchmark here.
Scaling almost exactly according to specifications, ps5 between 2070 to 2070S, xsx close to 2080 (even though he didnt test this with the new patch?).
Untill you start ray tracing ofcourse things would look different or go dlss vs the reconstruction used on consoles.
Game is supposedly gpu limited so we got a nice benchmark here.
AFAIK that's always been the way with Special Source abilities. Look good in the pre-release marketing and helps generate interest, but doesn't really change much in the final experience. With so few console exclusives, everything is made for a lowest, most vanilla common denominator, which excludes anything fancy. I think all you really get is a few more FPS when some bit of hardware can be used easily through an API to do some generic workload a bit quicker. But the tangible gains will just be a few FPS and never anything more.
I think he tested on both the new patch and the previous patch as the XSX is hitting the 40fps limit on the new patch and so its performance is being underrepresented which shows up as a 19.25% advantage to the 2080Ti. Looking at performance on the release version though where the XSX isn't limited it comes in at only a 16.2% deficit compared to the 2080Ti. Based on TPU's testing of the same game (albeit it in a different scene with different settings so not exactly comparable) which shows the 2080Ti performing about 30% faster than the 2080 then this looks like a pretty good showing for the Series X with it probably being a bit quicker than a 2080S here, although closer to the Super than it is to the 2080Ti
Yeah I agree. I am super disappointed as well. Not even Plague Tale, a current gen only game, supports these features. Not to mention Flight Simulator, which is a Xbox/PC only title.
Also you see close to no developers on Twitter even caring about features like Mesh Shader/SFS. And Mesh Shader support in UE5 is basically irrelevant if the r.NaniteMeshShaderRasterization command works as I expect it to. SFS is irrelevant too, because UE5 uses Virtual Texturing.
This is sad. We have consoles and PC GPUs with such cool features but nobody uses them, leaving a ton of potential performance and quality behind.
Right. Its still even for the XSX quite close to the indicated specs looking soley at that chart, would put it around 20% over the PS5, or closer to 6700XT, with the 6700XT somewhat quicker than the XSX most likely, however it also has 1TF more to compute with. The PS5 sits around 2070S, somewhat slower. And thats almost exactly 6600XT performance for the PS5, with the 6600XT having the slight advantage.
AMD Radeon RX 6700 XT Specs
AMD Navi 22, 2581 MHz, 2560 Cores, 160 TMUs, 64 ROPs, 12288 MB GDDR6, 2000 MHz, 192 bit
If anything, in this game the hardware is scaling very nicely with the given specs for consoles and pc GPU's. Its also GPU limited so a fair test for GPU performance comparisons across the board.
So while we can say its a good showing for the XSX, its right where it should be performing specwise. In the same vein, the PS5 is just as good a showing since its performing where it should be in this title.
Control was a good benchmarker, so now we have this one aswell.
For good measurement, its a good idea to have as close settings possible to each platform for fair comparisons. Altering settings can truly skew results.
Right, maybe we get 1-2 titles that are ambitious and try to push the hardware. Best thing that will happend we will see interesting GDC talk and thats it. Something exciting for tech enthusiats and nothing more.
snc
Veteran
Me too, sad df didnt do it also for uncharted 4 as there was also no dynamic res and uncapped mode so rather easly to compare
That’s just not simply true. It’s. very difficult to change the specifications of a game already inflight. It’s costly. Teams will leverage features as they require it. Many teams will have to test the features to learn the limitations before the “up” how much complexity the game can handle per frame.
Only certain teams in the first party space have had enough time, input, and testing to be able to leverage these features early on. Everyone comes in late. Most of these features may not come to use until closer to the end of the generation. It for certain studios, it may start coming within the next year or two. Considering how far Coalition has gotten, I think we can expect fairly significant integration on their next title. And if they release one following very deep integration.
Only certain teams in the first party space have had enough time, input, and testing to be able to leverage these features early on. Everyone comes in late. Most of these features may not come to use until closer to the end of the generation. It for certain studios, it may start coming within the next year or two. Considering how far Coalition has gotten, I think we can expect fairly significant integration on their next title. And if they release one following very deep integration.
Absolutely agree. I was saying even before these consoles launched that based on the specs the XSX should be close to the 2080Ti and the PS5 close to the 2080. What I think we've seen so far, outside of this and maybe a handful of other games (I've not done extensive research) is that the XSX tends to underperform a bit and the PS5 more often sits where we'd expect it to. I've no idea why and really don't want to start one of those debates up again, but it's nice to see things falling into place as we expect with this game, particularly with it being one of the first and only true cross platform next gen show pieces.
Me too, sad df didnt do it also for uncharted 4 as there was also no dynamic res and uncapped mode so rather easly to compare
Probably because it's a last gen game unlike Requiem which is one of the first next gen only cross platform titles. It should be a much better test of how PC GPU's can fare against the consoles in future titles than a straight up port of a game developed specifically for the PS4.
The team doing exclusive are most of the time the only one to fully push a hardware, This is always interesting. I hope we will have the same for Spiderman 2 or any other exclusive title or even for multiplatform UE 5 title.
snc
Veteran
Hard to say, so many engines, so many studios tough results almost exactly as hw specs shows months before consoles premiere ;d
DF has always noted that the PS5 would be closest to a 2070S, which seems what is happening (somewhat below that then). While the XSX sits around 2080S. Perfectly as they should be performing and what we generally see when theres no ray tracing involved.
I have to point out its generally more accurate to compare to AMD products since they share the same architecture and all that. Ballpark 6600XT and 6700XT. Also for more healthy discussions going forwards i guess.
Theres a reason why multiplatform titles are generally used to gauge performances across different platforms. It would be the same the other way around if something is ported to the PS5 at a later time.
I agree its intresting, however not in perspective of comparing performance between GPUs and platforms. Even though so far the ports that are quality, usually by Nixxess assistance, pc parts are generally performing where they should in relation to the PS5's specifications.
The features may not be very useful in practice. Geometry shaders and deferred contexts being examples.
I think that one was really useful, and some games ended up shipping with it, but since AMD did not support it a number of devs just never ended up adding it into their engine to boost DX11 performance on the CPU. I know particularly of one title from days of yore where a dev told me they had deferred contexts set up, but since AMD did not support, they did not ship to the detriment of NV user's CPU performance: "to keep things even".
Deferred contexts were a bad idea in general. Microsoft employees on the DirectX discord and others commonly recommends developers to not use this feature on D3D11. Also, using deferred contexts on Nvidia HW crashed Oxide Games Nitrous engine Star Swarm demo ...
Edit: Not that it matters but even Intel hated the idea ...
Last edited:
yea I would say it was the predecessor of where they wanted to go.
But if you can't code at a d3d12 level, then having some access to deferred context is still better than not on dx11 if that is what you required.
Metal Snake Eater
Newcomer
a 5700XT is around 9-12% faster than a non XT
as an example here is DF analysis of Death Stranding
the 5700 having 75% of the PS5 performance which is close to a Plague Tale Requiem of the Same GPU having 78% of the PS5 performance
while the 5700XT having 83% of the PS5 performance
so i am expecting the same gap in Performance in this Title as well
as Alex mentioned in the Video AMD GPU's doesn't perform that well on a Plague Tale Requiem compared to Nvidia as the 2070Super slightly outperforms the PS5 in this title while it under performs in Death Stranding despite the PS5 having the Same Performance advantage against it's RDNA1 brethren "Driver perhaps?"
Similar threads
- Locked
