• 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.

Next-Gen PS5 & XSX |OT| Console tEch threaD

Status
Not open for further replies.
I got warned for "Inappropriate behavior" half hour ago, with no link to a post.. anyone here know how to tell WTF you got warned for?

I'd ask somewhere else.. but.. there's no place I can tell to ask questions like that.

Have you posted in the Off-Topic section recently? That's probably a good place to start. Mods seem to be a bit more strict about moderation there than on the gaming side.
 

ethomaz

Banned
The RE I used to love was 3rd person.
View point doesn’t change the experience for me but a genre change yes.

People praises RE4 but I couldn’t play 30 minutes... the action genre and followed by the abomination RE5 and RE6 made me put the franchise on hold.

Thanks they back to original Survival Horror.

PS. To not say I never played another RE after 4... I have payed REmake on PS4 that is really great.
 
Last edited:

SlimySnake

Flashless at the Golden Globes
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
 
Last edited:
What's everyone's thought on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.

Well it could be correct that they focused more on the PS5 version than the XSX version.
 

SSfox

Member
What's everyone's thought on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.

PS5 is the definitive best hardware for gaming, not only the loading times are the best, and also the pad is the best, but also best in other technical aspects like fps et resolution.

Outside of the lack of full BC, Sony has done their job for the PS5 hardware, thanks to Cerny.
 

DeepEnigma

Gold Member
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
10 month dev cycle (under perfect circumstances), 5 new platforms to optimize for, and Covid with lockdowns in Cali while trying to work from home.
 

Thedtrain

Member
I just didn't know if there was a thread dedicated to such things.

But yes, this is our random thoughts thread lol
I got warned for random thoughts in this thread just an FYI.
Got to stick to next gen gaming or reply ban.

That fast travel is amazing though, glad the switch to SSD’s made it in this gen
 

Vagos48

Member
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
The explanation is simple. A SONY conspiracy!!
 
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
Based on the timing of the release the game was in development before the new consoles came out. The Xbox GDK was known to be immature and it has caused lots of developer issues hence all the complaints before release.

This has nothing to do with bottlenecks or Sony conspiracies but simply immature dev kit vs a mature one. Add in a Sony first party devs experience with a Sony platform and there you go a minor framerate advantage for the home team. I wouldn't read much into just like many don't read anything into the advantages shown in Hitman 3 on Xbox and PS5. Let's see where games are this time next year. Cross generational titles are not the best indicator of system power.
 
Last edited:

assurdum

Banned
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
He always said that. When there is a performance issue on Xbox, it's software relative hearing him. ps5 has some hardware advantage and sometimes could show it. Just depending of the engine. Series X GPU is superior in the compute calculation but not in everything. CPU is equal. About the bandwidth well the data management is very important so I guess ps5 could have a more efficient uses of it. Anyway it happens occasionally not every time.
 
Last edited:

SSfox

Member
He always said that. When there is a performance issue on Xbox, it's software relative hearing him. ps5 has some hardware advantage and sometimes could show it. Just depending of the engine. Series X GPU is superior in the compute calculation but not in everything. CPU is equal. About the bandwidth well the data management is very important so I guess ps5 could have a more efficient uses of it. Anyway it happens occasionally not every time.
Yeah i think we get it now on how those DF guys works:

If it's best on Xbox:

"As expected it run better on Xbox because it's the most powerful hardware with more teraflops"

If it's best on PS (which is the case most the times in multi games):

"The game is not optimized yet for Xbox"
"Hum, the Xbox version must have some bugs right there"
"So it seems Xbox dev tools are not ready yet"
"It's weird that it runs better PS5 than Xbox, i don't understand MUH"

ect ect...


Shouldn't even be surprising tho, it was already said by legit sources that they got their pocket filled with some MS dollars.
 
Last edited:

Fafalada

Fafracer forever
What's everyone's thoughts on this?
...
the CPU clockspeed is higher.
Just one tidbit here - but the clock delta between respective CPUs is practically irrelevant - it was like that last gen, and it only got smaller this time around.
Changing a single compiler optimization flag can account for larger disparity than clock delta between PS / XBox models, and then we have different APIs - where overhead can be much larger yet. So really - if there is a CPU difference between titles it'll usually be a hundred other things causing it before that mhz difference comes into play. You need at least 20-30% upclock before you'll start seeing meaningful differences (and that we occasionally do, in mid-gen refresh consoles).

That aside - the said sequences perform largely the same on last-gen consoles so it's highly unlikely there's CPU limitations in play, so whatever is going on looks to be rendering-throughput bound.
 

DaGwaphics

Member
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.

1X isn't as far ahead as it should be (definitely no 40% advantage) either. This is likely just a case of the game being built from the ground up for Sony apis and then ported to DirectX. The 1X can push ahead of PS4 Pro in spite of that because there is such a large difference performance between the two systems. The XSX and PS5 are obviously much closer with both trading a 15 - 20% advantage on opposite sides of the pipeline.
 

LiquidRex

Member
FidelityFX is a set of algorithms/tools available for developers using AMD hardware.... it is available to use since the hardware supports it.

Said that you need look at the names... MS announce and these being used in PS5 REVIII are different.

FidelityFX Contrast Adaptive Sharpening (CAS)
FidelityFX Single Pass Downsampler (SPD)
FidelityFX Combined Adaptive Compute Ambient Occlusion (CACAO)


This is interesting. 👍
 

Rea

Member
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
Cerny's secret magic sauce.
7vVZILB.gif
 

Iced Arcade

Member
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
Sony did a damn good job on the competitor's machine.


color me surprised it wasn't a noticeable difference. I would have put money down that a sony studio more familiar with its own hardware would have had a much different story. kudos to the San Diego Studio
 
Last edited:

Shmunter

Member
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
Yeah the Pro vs OneX and the PS5 vs XsX performance profiles are a head scratcher. You shouldn’t need to go out of your way to get decent results on the more modern console. I’m adamant it’s the 10/6 memory setup causing issues, and without proper care to optimise it’s letting the system down.
 

Vagos48

Member
1X isn't as far ahead as it should be (definitely no 40% advantage) either. This is likely just a case of the game being built from the ground up for Sony apis and then ported to DirectX. The 1X can push ahead of PS4 Pro in spite of that because there is such a large difference performance between the two systems. The XSX and PS5 are obviously much closer with both trading a 15 - 20% advantage on opposite sides of the pipeline.
40% more GPU TFLOPS alone doesn't translate to 40% more performance. Resolution is probably more affected by that, but fps depends on the CPU as well which is not that different between Pro and X1X. Maybe they could up the resolution a bit more on the X1X and even the frame rate.
 

OneTruth

Banned
Everything is important. OS, API, GPU drivers etc. But hardware is king, because you may improvement software, hardware - never.
 
Last edited:

Lysandros

Member
Why people are so willing to 'completely' ignore PS5's GPU advantages when a specific game/scene is running better in this machine is beyond me. It's like people are desperately wanting XSX to be more powerful in real gaming scenarios to reach some kind of inner peace, including quite a substantial number of Sony fans. That's hilarious.
 
Last edited:

ToadMan

Member
Why people are so willing to 'completely' ignore PS5's GPU advantages when a specific game/scene is running better in this machine is beyond me. It's like people are desperately wanting XSX to more powerful in real gaming scenarios to reach some kind of inner peace, including quite a substantial number of Sony fans. That's hilarious.

It's a hang up on the Teraflop concept and that itself is hangover from PC testing where the system is non-variable and just switch GPUs for the tests. No surprise the cards with more Tflops did better - of course they did, the test was designed only to test Tflop differences and tells you nothing about the performance of 2 different systems.

The same has happened in a lot of consumer electronics stuff. Remember when cameras were marketed only by their Megapixels while ignoring the quality of glass that got the light onto the CCD? Or when TV sales were reduced to terms like HD, or HD Ready and no one cared about contrast or lag or quality of input sources. And of course, when cars were all about HP, not to mention when processors were all about Clock speeds.

All those single point measures were incorrect - they became marketing terms because consumers didn't know enough to understand the tech and what it meant and often neither did the sales people and marketers. These were specialist items being sold to lay people and that meant dumbing things down for them.

And now you have console GPU Tflop comparisons throwing people for a loop. Unfortunately nothing in life can be reduced to a single number like that... its very inconvenient and not that many people have the motivation to educate themselves on the nuance - just click a like, nod in agreement and move on, even in the face of data that opposes the herd's views.
 

OneTruth

Banned
Why people are so willing to 'completely' ignore PS5's GPU advantages when a specific game/scene is running better in this machine is beyond me.

Because we dont have any real next-gen multiplatforms games. Old engines without any RNDA 2 effects and optimization. We must wait for PS5/XSX games only, to know more.
 
Last edited:

Elog

Member
I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.
It is consistent now across many titles so I have asked myself the same question. One thing that potentially could explain the difference is the difference in cache memory. The PS5 has both significantly more cache per CU (fewer CUs for the L2 cache and smaller arrays so more L1 cache per CU as well),and it is also faster. The higher up in the cache hierarchy you need to go to get information as a CU the longer it takes (L0>L1>L2>VRAM).
 

M1chl

Currently Gif and Meme Champion
What's everyone's thoughts on this?

hEmlT1J.png


This only happens in cutscenes when they up the graphics quality and use hero assets and better lighting. Richard said it's because the game is not yet optimized for xbox. And yet the xbox one x version outperforms the PS4 Pro like you'd expect it too.

I have brought this up several times now so this might seem like a broken record, but is there a bottleneck somewhere in the XSX? If so, where? The ram bandwidth is higher. the CPU clockspeed is higher. The GPU is more powerful. Clearly the DirectX vs whatever openGL libraries they are using for the PS4 versions are not the factor since the xbox one x version is outperforming the pro like it should.

If anything I wouldve expected the PS5 to drop frames like this based of all the pre-release hysteria over the variable clocks. People were convinced that the PS5 was actually 8-9 tflops because Cerny had said that the clocks would drop if the power consumption went above the power budget of the cooling system. Many bans were handed out, and I guess time has proven those bans legit because clearly this thing isnt 8-9 tflops and the 12 tflops console with fixed clocks is the one struggling to maintain performance in several games now.

I wish we had a PC release for this because then i could compare the performance against my rtx 2080 which MS said is equivalent to the xsx GPU based on their Gears 5 tests.
I just speaking as someone who talked to engineer of my past team (KC: D) and MS introduced shit ton of new APIs, like that you are basically rewriting games, not from scratch, but it's considerable effort, we didn't discuss any of the game performance, I was just curious about GDK. I was explained, that kind of have to do it, because GDK now can compile straight to windows binaries, so there is need for different APIs which then base on what you have in manifest do xbox or PC versions. However, you are not required to use them, sure Visual Studio will warn you, but you can compile last-gen API code by GDK without an issue. Which gives XSX version 12 TF of GCN (this expression sucks, but hopefully you get the point), because it does not use APIs for those RDNA features. Based on inner document this is going to get fixed, but it does not matter when since it's already too late and games moves to another engines.

Also you probably figured out, with old APIs the whole divided memory pool is fucked.

/rumor
But I guess it's not high-budget title, they are mainly PS devs, so why fuck around with new codebase, when they can change few things like resolution and effect and compile the old X1X codebase.

I am not saying that this is what happened, but could be.
\rumor

And again this is not really tools, but how MS handle the "generations"...
 

Lysandros

Member
It is consistent now across many titles so I have asked myself the same question. One thing that potentially could explain the difference is the difference in cache memory. The PS5 has both significantly more cache per CU (fewer CUs for the L2 cache and smaller arrays so more L1 cache per CU as well),and it is also faster. The higher up in the cache hierarchy you need to go to get information as a CU the longer it takes (L0>L1>L2>VRAM).
I also see the cache architecture and speed as one big contributor to the overall performance picture but of course not as the only one. I personaly don't think that XSX has 'a special and mysterious bottleneck' but it is less efficient and slower, thus perfoms more or less as it should.
 
Status
Not open for further replies.
Top Bottom