assurdum
Banned
I wouldn't take that occasional big advantage on ps5 as indication of something. With all respect.
Last edited:
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
DF- Devil May Cry 5 SE: PS5 vs Xbox Series X - The First Next-Gen Performance Face-Off
alabtrosMyster
Banned
That comparison never took any of the clock speed advantage into account... Everything else being equal the advantage would be around 20%, and the series X also has a memory bandwidth advantage (bit it seems at a loss in the caches for some reason).
Anyway, 44% was never to happen, except in the dreams of those who have Phil posters on their sleeping room wall.
ChuckeRearmed
Member
It seems after the last DDOS attack by sony fanboys, DF started to praise PS5 more
Like
Like
- BC runs at a lower resolution and has no enhancements -> PS5 has better BC.
- XBX runs in several mods better than PS5 and had some weird issues in another mode -> PS5 is nuts
Werewolfgrandma
Banned
It literally gets nearly the same fps at 4k as it does 1080p. Doesn't that strike you as odd? The most logical conclusion is something isn't working right. To suggest anything else is silly.
assurdum
Banned
I understand the feeling. I hate when they lead such narrative just to the MS side. That's annoying and not fair.
Last edited:
ethomaz
Banned
I believe the goal of the video is a graphic/performance comparison.
The Haptic Feedback doesn't affect that.
Last edited:
alabtrosMyster
Banned
You mean it's showing by being barely above... Not even close to the promised 20%.
This is a huge hype killer for Xbox as "power" was their only advantage... Unless their studios start shipping really amazing games FAST.
NoMoChokeSJ
Banned
It isnt broken because it didn't win... you have a sometimes 40 fps variation during the same scene, same fight, same enclosed environment.... nothing changing. Thats extreme. The Ps5 is wildly all over the place in the mode as well, but not as dramatic.
Truespeed
Member
Yep, I was told that XSX would have 44% more perf in RT mode since XSX has 44% more RT Cores in CUs.
Reality set in when they put it on the dyno.
Romulus
Member
It's not and there's no like for like comparisons between them either.
Its only working "right" when XSX isn't "winning."
Romulus
Member
It's just not like for like, different enemies onscreen, camera close vs far away. I can jump 40fps on an unlocked game easily doing that.
Last edited:
Caio
Member
From min 06:00 , high frame rate mode, PS5 can keep 120 fps while XSX drops at 75 fps, and I found this really weird... API limitations ? Looks like the GPU is being held back by "something", which cannot be CPU or memory bandwidth, of course !
Romulus
Member
Yeah I mean that's another argument altogether that can be made. So, all modes are broken I guess because its not showing anywhere near the power advantage, not even remotely close.
Werewolfgrandma
Banned
What are you taking about? I couldn't give 2 shits about Xbox. It's just logical and you are just making stuff up.
MasterCornholio
Member
Well it's true that the API could cause some issues but it's only temporary. Let's see what happens later in the gen. He can't spend the whole generation making the claim if the issue persists years later.
Romulus
Member
i didn't say you did care about Xbox. I'm addressing my original point that you're attacking. Be specific about what I'm making up.
Lysandros
Member
It was extremely naive to believe that a GPU with only 18% more vector ALU throughput while having a 20% deficit in nearly everything else (pixel fill rate, rasterization/polygon throughput, cache bandwidth, async compute, GE/culling ect.) would perform 18% better anyway.
Last edited:
Michael Hawk
Member
I guess we're just going to ignore that DMC5 was built for last-gen consoles.
Truespeed
Member
You're not factoring in the API that produces that 18% theoretical advantage - which is apparently broken according to sources familiar with the matter.
Self
Member
No stable framrate = shit.
alabtrosMyster
Banned
I know, I had that argument with someone else in a different thread.
SlimySnake
Flashless at the Golden Globes
goddamn. this guy called it months ago.
alabtrosMyster
Banned
Equals then....
Werewolfgrandma
Banned
Me not caring about Xbox is directly about your "point"
To come to any other conclusion then this mode isn't working right is silly. Anything else seems about as reasonable as saying it's sea monkeys is the culprit.
geordiemp
Member
What sources ? Its all gossip ?
BlueXImpulse
Banned
If we see more examples of this in practice it'll be easy to conclude that he was spot on.
geordiemp
Member
Well he would know what ps5 has done here.
Truespeed
Member
MasterCornholio
Member
I just want to make sure I understand this CCX thing.
The reason that one cluster is better than two is because of the shared cache. Which is essentially something that Zen 3 is doing with it's CPU clusters.
Is this correct?
Last edited:
ethomaz
Banned
Basically yes.I just want to make sure I understand this CCX thing.
The reason that one cluster is better than two is because of the shared cache. Which is essentially something that Zen 3 is doing with it's CPU clusters.
Is this correct?
It is where all the rumors about using Zen3 tech come from.
BlueXImpulse
Banned
I just want to make sure I understand this CCX thing.
The reason that one cluster is better than two is because of the shared cache. Which is essentially something that Zen 3 is doing with it's CPU clusters.
Is this correct?
Precisely, you want to minimize latency whenever possible.
JerryinSoCal
Member
Blame the devs for saying it wouldn't.
MasterCornholio
Member
The good thing about this is that we will see it in a die shot. Should be easy to prove.
ethomaz
Banned
X-ray yes... just a die shot I believe not being possible to see that.
NoMoChokeSJ
Banned
I'm not talking about a comoarison between the two. There are times the xbox drops 40isj frames during a fight with a single enemy in an enclosed environment, then jump up 20 frames. Its all over the place in a relelatively static situation.
Romulus
Member
Then you're not reading my point. I made it clear the 4k and 1080p modes don't show like for like with each other, lots of disparities between the shots with camera rotation and enemy counts. Again, I can make unlocked framerate games spike well over 100fps in games that normally do 60fps.
Now, should both consoles run at a higher framerate in performance mode? Probably, but we don't know the full story. The XSX and PS5 operate at a very similar framerate in performance mode. There's no consistent huge gap where PS5 is just constantly pulling away, its close. I just found it odd that only the XSX gets a pass as possibly "broken" in this mode. Just makes no sense to me with all the other unknown and inconsistencies to have any conclusion really. It could very well be that PS5 or Xbox are very unoptimized, but some would hate to entertain that PS5 is unoptimized because it takes away from where it wins in this comparison. There's this narrative that "ps5 is being maxed out Japan devs!" Wtf
Last edited:
Romulus
Member
John pointed that out, and even made mention "this one area is strange with XSX." Meaning, its not often.
Silent Viper
Banned
....Bad optimised ports
PaintTinJr
Member
I'm only just into the video, but if they are really claiming the picture on the right has the same rendering setup as the one on the left, then it is no wonder they incorrectly used the word "Frustrum", when they mean "Frustum" in the article text.
In this shot, it shows the PS5 has a wider field of vision and a further draw distance - because the backgrounds are perfectly aligned, but on the PS5 you can see far more of the foreground at his feat, and on the XsX the building on the right is showing another section of glass, because the near/far clip plane relationship is closer. You can also see that the PS5 is showing more balloons and specular highlights - not sure if reflections, but they probably are on each ballon - whereas the balloons on XsX look very flat, like VRS has only fully shaded the full front facing fragments.
It is the same comparison with AC. It is hard to see the differences, but both systems are doing different workloads, and in both cases the PS5 is doing more. So this is a bad start for DF IMHO, but obviously I'm only part way through the video
assurdum
Banned
Of course
Silent Viper
Banned
DMc5 new version sucks
Weird how they talk about the game. This exact frame isn’t a great comp, but if you look at the graph Xbox doesn’t dip below 60 in the smoke while ps5 dips into the 40s. You’d think that would deserve a mention
Lysandros
Member
CPU can be 'somewhat' responsible for it due to specific customizations to reduce cache latency (unified L3?) as hinted recently by Matt Hargett, the principal software engineer of PS5. Especially considering low resolution/(very) high framerate of the performance mode. This mode is more likely to be CPU bound i think. There is also the matter of rasterization which doesn't scale with resolution and PS5 has a nice 20% advantage there.
Last edited:
NoMoChokeSJ
Banned
The Japan thing is dumb... moving past that. The idea that the other modes aren't broken on the PS5 is silly to me as well. They run consistently, without the wild swings in fps that the high framerate mode experiences. The biggest slow downs you see on the PS5 are in effects heavy areas during the 4k footage. If they were experiencing wild changes in FPS without environment changes, or if the swings happened in the more demanding modes, I would agree the game is unoptomized as a whole, but that's not what you see in the video. You are arguing about two different things imo, optimization and a technical bug.
Truespeed
Member
We've moved beyond that. We're now into conspiracy theories.
Similar threads
- Locked