This post is not entirely accurate tho, performance does matter but only to a certain degree. PS4 was quite substantially faster at around 40% and the resolution gate was especially prevalent at the time when temporal reprojection was non existent. This time around the power difference is much smaller at around 18%, not to mention a 2.2x SSD advantage to PS5, both machines should easily do 4k, raytracing, 60fps. With the smart usage of dynamic res and temporal reprojection, the difference would be further smaller especially when talking about 2000p vs 2160p. Price is somewhat more influencing tho, but hell even if PS5 is $50-$100 more expensive the SeriesX it would still be the much faster selling console, the brand alone and the exclusives are just that much more popular these days. Perhaps things would change after tonight's showing but for now I stand by my points.
That RDNA 1.8 Consoles Rumor *spawn*
- Thread starter Ronaldo8
- Start date
If flops was the deciding factor there would only be pc master race left. Sony very well could have vrs as implemented by amd. Isn't the whole beef in microsoft implementing some sort of better vrs similar to how sampler feedback makes streaming potentially more efficient.
For sampler feedback one big question is that is collecting the misses and fetching pages after the fact only possible solution? What if the missed pages are not needed anymore once they have been fetched? What if some other heuristics or even machine learning could be implemented to similar effect without sampler feedback and avoid the misses to begin with? In this case running something like sampler feedback would help on training ml algorithm but the actual runtime network would not need sampler feedback.
edit. What I'm trying to say that proof will be in the pudding or in this case games. It's not wise to predict doom&gloom based on paper specs. My suspicion is that it will all come down to art and not to flops. i.e. who has most talented development teams and most money/time to allow those teams to achieve their vision.
For sampler feedback one big question is that is collecting the misses and fetching pages after the fact only possible solution? What if the missed pages are not needed anymore once they have been fetched? What if some other heuristics or even machine learning could be implemented to similar effect without sampler feedback and avoid the misses to begin with? In this case running something like sampler feedback would help on training ml algorithm but the actual runtime network would not need sampler feedback.
edit. What I'm trying to say that proof will be in the pudding or in this case games. It's not wise to predict doom&gloom based on paper specs. My suspicion is that it will all come down to art and not to flops. i.e. who has most talented development teams and most money/time to allow those teams to achieve their vision.
Last edited:
The proof will be in the pudding as they say.
We have yet to see the differences between consoles in games. One console can still hold an obvious advantage over the other. Or most people wont be able to notice.
We will have to see when we get the boxes in our hands. But to the point of console transitions. Its not hard to go on psn or xbox live and tell your friends hey this is my name on the other platform lets all move to that with the new consoles because this game series we have been playing for years will run better on it.
I think long term multiplayer games have better lock in. Think sea of theives. THere are people who have played it since it came out and enjoy the regular content up dates. It will be a lot harder to convince someone to switch platforms with that , then someone who might have liked a single player game and could be waiting years for a sequel. I heard that ghost game on ps4 that just came out is great. but if you want to play the sequel on ps5 how many years do you have to wait ? 2 ? 3 ? In the mean time what if the other games you play are cod or destiny or some other big multiplatform game. If one console offers a higher resolution or better frame rates you might go for that. It becomes more interesting when you add in mid cycle refreshes. IF you know that single player sequel wont be out till around the mid gen refresh would you buy the more powerful console and then wait for the refresh and that single player game to grab it.
Lets also not forget to factor in the cost of tvs. Vizio has a new line up of tvs that start at $230 and have hdr at 40 inches. Its going to be cheaper to get a 4k tv than it will be the console at launch. even slight frame rate advantages could push people over the edge as freesync is coming to more and more tvs also
Really? 20 % more GPUflops plus Variable Rate Shading easily = 50-60% difference in real term. And the RDNA2 Ray-tracing method use the CUs. More CU = more RT power, Xbox has 36 + 16, PS only 28 + 8 = 100% advantage. Theoretically, maybe more.
D
Deleted member 86764
Guest
Not just that the post doesn’t even consider one was out before the other and was significantly cheaper (core/arcade) and had a tonne of great exclusives vs a floundering expensive opponent.
Also launch line up is not the be all, you know with each company what you will get roughly (exclusives wise) you don’t buy based on the launch line up.
HBRU
Regular
Indeed. Looking back over the launch titles, there's really not that many exceptional titles, especially for the past PS4/XO generation. There were far better launch titles for the previous generations -- https://forum.beyond3d.com/threads/console-launch-titles-reference.61730/
HBRU
Regular
No. VRS won't ever be that much of a multiplicative impact the first couple years. If it's ever to increase beyond an additional boost range of 10-20% (forget exact increase from the Gears Tactics Articles), it will require more research and development into new algorithms and procedures, where VRS would be applied outside just the final shading pass.
Rootax
Veteran
Not me. + I don't like the weird bandwidth distribution on xbox.
So the XBSX will be 160%+ faster than PS5 in ray traced games ?? Sounds like Sony is doomed and should get out now before they embarrass themselves.
Silly Sony
A 12TF+ TF gpu, faster clocked CPU, much more bandwith to the GDDR ram. Aside from that SSD tech will be close, MS does efficiency sony goes for raw speed.
D
Deleted member 7537
Guest
RDNA 1.9.1 beta still on the lead. IMO of course.
Gears Tactics was only tier 1 - tier 2 in RDNA2/DX12U should be a lot better.
Plus, QPlayer was stacking the TFLOPs advantage with VRS. Even if it’s only 20%, those stack to more than 40% combined. I think it could be close to 50% at the end of the day.
umm. VRS 2.0 is max 15% boost lol.
Last edited:
I realize it’s an over simplification, but in principal a performance advantage would stack with an efficiency improvement.
1.18 (raw performance advantage) x 1.2 (reasonable estimate for VRS efficiency improvement) = 1.42
We really don’t know what tier 2 VRS will do. Only tier 1 lines up with what you’re saying.
