Support NeoGAF

[Neko Mimi Mode]

This is interesting:

~30% better performance here on XSX, GPU bound scenario (cutscene). But we don't know how dynamic resolution system works, is XSX the same res, higher or lower?

Presumably the dynamic resolution scaling works like any dynamic resolution scaling - it drops the resolution up to a predetermined minimum to maximize frame rate in scenes that demand it.

Here, because PS5 is dropping pretty consistently into the 40s, it's safe to suggest that it's hitting said minimum and can go no lower. Almost certainly XsX will be at the same resolution or higher.

It could be that XsX has lower dynamic resolution minimum, but if the game was dropping to the 40s on PS5, they'd have given it the same too.

Someone is arguing the different scene in the comparison, you don't get the point. The point is there is no such big drop on PS5

It's not even that it's a different scene, it's a different mode entirely.

XsX gets drops on killcam/blackscreen on online MP, but we don't know if PS5 gets the same. The PS5 video is an offline private match against bots.

 

[SLB1904]

These comparisons aren't interesting anymore. Who would have thought tf aren't everything... oh nvm cerny knew

 

[geordiemp]

And people were talking about damage control in the beginning of this thread, delicious.

Again, I think this thread will now quickly die considering the expected result isn't exactly obtained.

One more thing: dropping from 120 fps to (let's take the lowest) 82 fps is something you will barely notice. But going to 23fps in a cutscene... that will hurt.

That cut scene 30 FPS analaysis was just wrong, not too sure what they are using but doe snot seem to work so well.

 

[assurdum]

And people were talking about damage control in the beginning of this thread, delicious.

Again, I think this thread will now quickly die considering the expected result isn't exactly obtained.

One more thing: dropping from 120 fps to (let's take the lowest) 82 fps is something you will barely notice. But going to 23fps in a cutscene... that will hurt.

Oh definitely I see damage control in this thread. Especially in the video posted.

 

[Bo_Hazem]

Not to sure the source, but does not seem to be even counting the 30 FPS cut scenes correctly.

Yes, cutscenes are 30fps but that app seems to have wacky algorithm.

 

[FranXico]

Instead we've got XsX online and PS5 against bots.

You mean the PS5 had the CPU a bit more taxed with the bots AI?

 

[assurdum]

Presumably the dynamic resolution scaling works like any dynamic resolution scaling - it drops the resolution up to a predetermined minimum to maximize frame rate in scenes that demand it.

Here, because PS5 is dropping pretty consistently into the 40s, it's safe to suggest that it's hitting said minimum and can go no lower. Almost certainly XsX will be at the same resolution or higher.

It could be that XsX has lower dynamic resolution minimum, but if the game was dropping to the 40s (or low 20s) on PS5, they'd have given it the same too.



It's not even that it's a different scene, it's a different mode entirely.

XsX gets drops on killcam/blackscreen on online MP, but we don't know if PS5 gets the same. The PS5 video is an offline private match against bots.

Almost certainly you don't know it, with all respect.

 

[Agnostic2020]

And people were talking about damage control in the beginning of this thread, delicious.

Again, I think this thread will now quickly die considering the expected result isn't exactly obtained.

One more thing: dropping from 120 fps to (let's take the lowest) 82 fps is something you will barely notice. But going to 23fps in a cutscene... that will hurt.

Lol talking about fall from grace . 2 weeks ago xsx was gonna crush ps5 now cgi cutscene 4 fps better on xsx but frames being worse during gameplay than ps5 is considered impressive for xsx .

 

[TheKratos]

Becoming clearer and clearer that devs have been able to work with the Ps5 longer and better than they have XSX. The XBox will catch up over time, but will it be too late in terms of market share? If CoD is perceived to run better on Ps5, that’s a lot of sales going Sony’s way.

COD was already more popular on Sony platform (hence the stupid exclusive deals) so a couple of frames more on either SX or PS5 won't matter at all.

 

[Armorian]

And people were talking about damage control in the beginning of this thread, delicious.

Again, I think this thread will now quickly die considering the expected result isn't exactly obtained.

One more thing: dropping from 120 fps to (let's take the lowest) 82 fps is something you will barely notice. But going to 23fps in a cutscene... that will hurt.

You mean dropping framerate in gameplay is NOTHING but dropping in CINEMATIC/NON PLAYABLE cutscene is a problem?

Also where you get 23 number? It's still only one frame lover than movies...

 

[Neko Mimi Mode]

You mean the PS5 had the CPU a bit more taxed with the bots AI?

I mean it's not a direct comparison, no matter which side benefits from that, it's apples to oranges.

Almost certainly you don't know it, with all respect.

Fair enough, so we wait for DF then? Agreed?

 

[geordiemp]

Lol talking about fall from grace . 2 weeks ago xsx was gonna crush ps5 now cgi cutscene 4 fps better on xsx but frames being worse during gameplay than ps5 is considered impressive for xsx .

Its also showing 0 and lots of other stuff for a 30 FPS cut scene, it is poor analysis software from an unknown channel.

But its funny seeing some excited about that poor cut scene video - let them believe

 

[GHG]

Two systems that achieve similar results by different means.

There's more than one way to skin a cat.

 

[Physiognomonics]

In DMC5 XSX already had slight advantage in 60fps gameplay, but as the game already runs at ~60fps on mid-gen consoles, the game is mainly GPU bound at 60fps, allowing XSX to have a small advantage (3% to 8%). But once in the 120fps mode, the CPU is now heavily taxed and suddenly the PS5 has the advantage.

Almost everything we have seen so far show that XSX, compared to PS5, has a CPU bottleneck. Maybe RGT is right when he says PS5 Zen 2 CPU has a unified L3 cache like Zen 3. That would explain everything (and no need for immature vs mature API narrative).

 

[Snowdonhoffen]

Not to sure the source, but does not seem to be even counting the 30 FPS cut scenes correctly.

Yes, those are literally video files so it's really weird.

 

[geordiemp]

Yes, those are literally video files so it's really weird.

Its not weird its just wrong, but the source is unknown so who cares.

 

[Agnostic2020]

In DMC5 XSX already had slight advantage in 60fps gameplay, but as the game already runs at ~60fps on mid-gen consoles, the game is mainly GPU bound at 60fps, allowing XSX to have a small advantage (3% to 8%). But once in the 120fps mode, the CPU is now heavily taxed and suddenly the PS5 has the advantage.

Almost everything we have seen so far show that XSX, compared to PS5, has a CPU bottleneck. Maybe RGT is right when he says PS5 Zen 2 CPU has a unified L3 cache like Zen 3. That would explain everything (and no need for immature vs mature API narrative).

Maybe the io of ps5 really unloads the cpu (cpu doesn't get involved ) and makes it perform better , while on xsx cpu is involved with io tasks

 

[skneogaf]

I'll wait for digital foundry.

 

[Stilton Disco]

So three for three showing basically identical results, with wins here and there on both sides, and a very minor overall win for PS5 that's barely anything to right home about.

This is seriously marvellous. Clearly the odd approach Sony took isn't holding things back, and thus both systems are delivering fantastic results, which is exactly what you want given they share an identical price tag.

 

[Werewolfgrandma]

Wouldn't be surprised if sxs has very slight adv during gameplay. Like .1 or less. Ignoring killcam drops. Could very well be to do with offline vs online.

 

[assurdum]

I mean it's not a direct comparison, no matter which side benefits from that, it's apples to oranges.



Fair enough, so we wait for DF then? Agreed?

Honestly, I don't even care. But would be fun if series X will "win" (and probably in the same subtle way as always), and the narrative around the mystic power in the series X hardware to rise again.

 

[Great Hair]

Essentially the same during live gameplay. Only during replay is where the X drops down to 80s.

 

[sircaw]

Is that PARITY i see.

 

[SoraNoKuni]

It gets even better, now it's an API problem lel.
I thought DX12 Ultimate pro extreme was a godsent.

I get they didn't lie about the velocity architecture, it's dropping frames in high velocity.

 

[captainraincoat]

Most people forget the console launch post patches on games.....gonna be a fun few weeks

 

[Agnostic2020]

It gets even better, now it's an API problem lel.
I thought DX12 Ultimate pro extreme was a godsent.

True ,To the point when ps5 rdna2 doesn't use dx12 U,it becomes an incomplete rdna2 as per those few hard-core fans .

I guess thank god ps5 is incomplete rdna2 ?

 

[MasterCornholio]

In curious to see the results in Watchdogs. As for CoD I'm going to wait for more comparisons.

I'm pretty much confused with what's going on.

 

[assurdum]

Most people forget the console launch post patches on games.....gonna be a fun few weeks

And what should happen exactly when post patches will be released?

 

[sinnergy]

And what should happen exactly when post patches will be released?

Performance degradation for both , and removal of RT, of course.... what do you think? Because that’s the goal of patches that we had for games the pas 15 years ...

 

[Agnostic2020]

In curious to see the results in Watchdogs. As for CoD I'm going to wait for more comparisons.

I'm pretty much confused with what's going on.

To put it simple in cpu bound modes ps5 is outperforming xsx and in gpu bound modes xsx is getting 5 to 8% better results .

Atleast thats my conclusion till now

 

[assurdum]

Performance degradation for both , and removal of RT, of course.... what do you think?

Not necessarly.

 

[Insane Metal]

* Lazy devs
* Just a launch title
* Bad tools
* Marketing agreement with PS5

etc

 

[yurinka]

Seem pretty much the same 99% of the time, with sometimes a small lead for one or sometimes for the other. In the campaign cinematics sometimes when it drops there is a XSX bigger lead. But well, I'd wait for more comparisions from NX Gamer or DF.

Do we know the loading times differences? And how the activity cards work on PS5 for this game? I assume they will be a quick access to different game modes or campaign spots to load them directly skipping menu and logo screens.

True ,To the point when ps5 rdna2 doesn't use dx12 U,it becomes an incomplete rdna2 as per those few hard-core fans .

I guess thank god ps5 is incomplete rdna2 ?

Pretty good job for a RDNA 1.5 9tf machine with no 120fps and only 100 BC games and a 40% inferior hardware.

 

[Agnostic2020]

* Lazy devs
* Just a launch title
* Bad tools
* Marketing agreement with PS5

etc

Compare that to 2 weeks ago when ps5 was gonna get crushed and it would downclock non stop and gap would be upwards of 50% . Wonder what changed

 

[MasterCornholio]

To put it simple in cpu bound modes ps5 is outperforming xsx and in gpu bound modes xsx is getting 5 to 8% better results .

Atleast thats my conclusion till now

Definitely not what I was expecting from these two systems. I thought the XSX would have a massive lead from the way that Microsoft marketed it.

Just seems weird to me.

 

[sircaw]

Definitely not what I was expecting from these two systems. I thought the XSX would have a massive lead from the way that Microsoft marketed it.

Just seems weird to me.

4 words

NEVER TRUST PHIL SPENCER.

 

[Bo_Hazem]

XSX is performing better here

By the way, cutscenes are 30fps. So this app isn't counting it right.

 

[GHG]

Most people forget the console launch post patches on games.....gonna be a fun few weeks

It would better for people to accept the reality.

 

[Werewolfgrandma]

To put it simple in cpu bound modes ps5 is outperforming xsx and in gpu bound modes xsx is getting 5 to 8% better results .

Atleast thats my conclusion till now

Killcam drops in cod have been an occurrence for a long time. Not sure the issue but on PC cpu and GPU utilization drops for a moment for killcam. Xsx holding 120 during gameplay maybe slightly better then ps5 so I'm not sure your conclusion is correct.

 

[Chukhopops]

Lol talking about fall from grace . 2 weeks ago xsx was gonna crush ps5 now cgi cutscene 4 fps better on xsx but frames being worse during gameplay than ps5 is considered impressive for xsx .

Ah yes, I forgot the rules of the comparison threads:

- PS5 performs better = PS5 wins;
- Parity = PS5 wins for some reason;
- XSX performs better, but not by some arbitrary huge percentage = PS5 also wins.

The goal posts have been moved so far that there is no way at this point for any comparison to be won either way, it's just pages of circlejerking like the AC thread.

But the best part is that the MP comparing different modes in different scenes is considered 100% gospel while a straight, scene-to-scene SP comparison is extremely questionable for some reason.

I would rather have 120 fps dropping to 83 during action than 60 dropping to 45, that is an obvious thing that shouldn't even be up for discussion.

 

[Elios83]

Definitely not what I was expecting from these two systems. I thought the XSX would have a massive lead from the way that Microsoft marketed it.

Just seems weird to me.

A massive lead was not to be expected anyway since both systems are pretty much the same with an on-paper 18% difference on the TFs figure alone and a specific figure doesn't tell the whole story about real world performance of a computing system. nVidia is claiming astronomic TFs figures for their 30xx lines of GPUs compared to their previous 20xx line but the actual performance differences are not what the TF difference would suggest.
But these results are fascinating because PS5 is actually having small advantages.
Imo it would be nice to understand what's going on on the CPU side and if certain design choices made by MS like the split memory with two different speeds for GPU and CPU is creating a bottleneck.

 

[Neko Mimi Mode]

XSX is performing better here

From the comments of this video...

Bryank75 this you mate?

 

[Werewolfgrandma]

Ah yes, I forgot the rules of the comparison threads:

- PS5 performs better = PS5 wins;
- Parity = PS5 wins for some reason;
- XSX performs better, but not by some arbitrary huge percentage = PS5 also wins.

The goal posts have been moved so far that there is no way at this point for any comparison to be won either way, it's just pages of circlejerking like the AC thread.

But the best part is that the MP comparing different modes in different scenes is considered 100% gospel while a straight, scene-to-scene SP comparison is extremely questionable for some reason.

I would rather have 120 fps dropping to 83 during action than 60 dropping to 45, that is an obvious thing that shouldn't even be up for discussion.

It's not dropping to anything lower then 119 during action. It's literally only the killcam and first half second of respawn that isn't basically locked 120 on xsx.

 

[Monkeygourmet]

 

[MasterCornholio]

A massive lead was not to be expected anyway since both systems are pretty much the same with an on-paper 18% difference on the TFs figure alone and a specific figure doesn't tell the whole story about real world performance of a computing system. nVidia is claiming astronomic TFs figures for their 30xx lines of GPUs compared to their previous 20xx line but the actual performance differences are not what the TF difference would suggest.
But these results are fascinating because PS5 is actually having small advantages.
Imo it would be nice to understand what's going on on the CPU side and if certain design choices made by MS like the split memory with two different speeds for GPU and CPU is creating a bottleneck.

In also wondering if Sony has more hardware that takes the load off their CPU than Microsoft does?

Like some of the customizations made to the I/O for example.

These results were definitely not what I was expecting.

 

[sircaw]

Definitely not what I was expecting from these two systems. I thought the XSX would have a massive lead from the way that Microsoft marketed it.

Just seems weird to me.

The thing is, the vast majority of the gamers out there, super casual people have no idea what any of this means, they need to be guided in a certain direction. Microsoft is appealing to them with their messaging and i am sure it works rather well. They are masters of pr spin.

This site however is rather hardcore and knows bullshit when they see it.

Even the hardcore xbox fanatics know this, but go along with it because it's their narrative.

Xbox is still a good machine, and even if it comes out ahead of ps5 i think the ps5 is better still, its got to many great extras for normal people.

Xbox fucked up this start big time, they had 7 years to prepare, they get no sympathy from me.

 

[Agnostic2020]

Ah yes, I forgot the rules of the comparison threads:

- PS5 performs better = PS5 wins;
- Parity = PS5 wins for some reason;
- XSX performs better, but not by some arbitrary huge percentage = PS5 also wins.

The goal posts have been moved so far that there is no way at this point for any comparison to be won either way, it's just pages of circlejerking like the AC thread.

But the best part is that the MP comparing different modes in different scenes is considered 100% gospel while a straight, scene-to-scene SP comparison is extremely questionable for some reason.

I would rather have 120 fps dropping to 83 during action than 60 dropping to 45, that is an obvious thing that shouldn't even be up for discussion.

Whatever makes u sleep better at night bud . I m extremely happy with ps5 results though .

 

[captainraincoat]

And what should happen exactly when post patches will be released?

Usually a big improvement across the board

 

[MasterCornholio]

Usually a big improvement across the board

So for PS5 as well? It's the first time that I'm hearing this.

 

[DrumNBassDanTTV]

Comments like this are getting boring and don't help the discussion unless of course your intention is to fuel the war.

Unlike the OP`s first comment?