• 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.

Lysandros

Member
FMXIAfP.jpg


Not related to RT this time, just (slight) graphical detail.
 
The only problem i can foresee is that the framedrops are way too frequent and unpredictable even on PC. My 2080 was constantly flactuating between 43-54 fps during any kind of physics activity. I didnt even have to be in combat for it to dip to 43 using PS5 settings. Just lift up something and throw it at something else and the framerate goes all over the place. It was a locked 60 fps when standing still which is very close to what we are seeing here on the xbox version.

In the corridor of doom, it goes all the way down to 33 fps. We have seen it dip below 30 fps in DF's gameplay recordings. But who knows maybe a 1080p mode might not have those drops.
Control isn't a very cpu bound game. In fact, it looks to be leaning more to the gpu side with all the particle effects, volumetrics, caustics etc. It's very taxing on the Gpu as evidenced by the graphics literally causing even rtx 30 series gpus to choke with all that eye candy going on when maxed out and at resolutions above 1080p. I don't expect consoles to do any better when the most powerful box is on par with a 2070s on anything but RT. And RT itself doesn't just require dedicated hardware for more efficient acceleration of the bvh structure, the structure itself requires high amounts of vram bandwidth.
 

SlimySnake

Flashless at the Golden Globes
Do you drive your cars by having them run on idle and looking at the RPM meter, or do you you actually drive them around town? Because car 1 is the Xbox, getting 5000rpm on idle, and then not going past 2500rpm when being driven, and car 2 is the PS5, that get's 2500 rpm on idle and 5000rpm while being driven.

Since cars are to be driven, congrats, Alex just proved that the shitbox runs better while not being driven around.
Again, the analogy doesn't apply here because the games engine or the gpu isn't idle. You can see gpu usage in games when standing still and it still operates at 80-90%. The car engine doesn't do that when you are idle.

Regardless, this is just a cool little experiment that gives us a direct comparison between the two gpus. No one is claiming that this is in game performance. It's no different than the cutscene comparisons Alex did with cod and ac.
Because you can't actually buy a PS5 anywhere. I could get the XSX within 5 days, PS5 is nowhere to be had.
Right. The Xbox one actually outsold the ps4 in December 2013 a month after launch.
 
Last edited:

LucidFlux

Member
aNOTp7i.jpg


Interesting..

Edit: Taken form beyond3d, don't know the main source.

Very interesting. Now it's a very small difference but the PS5 character is slightly closer to the edge of the steps. I wonder if this could be right on the edge of the cutoff of whether or not that particular object has its reflection calculated and the XbsX character is just a step too far back. Just an observation and a possibility.


FMXIAfP.jpg


Not related to RT this time, just (slight) graphical detail.

Now it could be how these images are cropped but notice how the camera on PS5 again looks slightly further forward which could affect which LOD levels are called. Specifically for that texture on the right. The slight difference in distance could even explain the fog and how much we can see of the what appears to be a manhole cover.

If both consoles truly share the exact same settings I can't think of any other reason why the scenes would differ.
 

assurdum

Banned
FMXIAfP.jpg


Not related to RT this time, just (slight) graphical detail.
The circle in the right seems indicate a bit better AF or LOD? Hard to say. The other 2 are practically nothing. One is the fog which is animated dynamically so just a different moment on series X, the other detail missed indeed is just the different perspective which trick a bit but the manhole cover is there too hide by the water.
 
Last edited:

LucidFlux

Member
Remedy themselves confirmed the Control settings are identical, people are trying to claim Remedy are wrong?😂😂

Well unless my eyes are lying something is certainly... different between the two versions.

I haven't had time to go back through the DF video to look for any of these differences in motion, but these screens posted lead me to believe it could be a matter of camera distance affecting LODs and reflection cutoffs.

I'm still giving Remedy the benefit of the doubt when they said the settings are identical, but surely you can see with your own eyes that something is off?
 

Rudius

Member
Any word on Mafia Definitive
Edition getting a next gen patch?

I played a bit of the old one years ago and would like to buy this, but a remake released so late in the previous generation should at least get an option for 60fps.
 
Last edited:

FrankWza

Member
Because eBay pricing is the defacto trusted source to indicate demand?
In this scenario? Absolutely. Are you going to pretend it isn’t? There’s a lockdown and one has retail units available and the other is sold out. eBay sales don’t count as units sold during that timeframe because they were snatched up when they were in stores, probably a month or 2 prior. Come on buddy
 

Thirty7ven

Banned
Because eBay pricing is the defacto trusted source to indicate demand?

It’s basic economics, if you don’t understand how supply and demand determines prices... you’re either too young to have learned or you skipped school and it shows. Being ignorant isn’t cool man.

By the way, as far as UK goes, for 2020, all Xbox consoles were behind Switch, PS5, PS4. In that order.
 

Rudius

Member
Well unless my eyes are lying something is certainly... different between the two versions.

I haven't had time to go back through the DF video to look for any of these differences in motion, but these screens posted lead me to believe it could be a matter of camera distance affecting LODs and reflection cutoffs.

I'm still giving Remedy the benefit of the doubt when they said the settings are identical, but surely you can see with your own eyes that something is off?
"Are you going to believe me or your own eyes!?"
 

SlimySnake

Flashless at the Golden Globes
I am sure Slimy is a very nice fellow but his posts can often result in such psychological alterations, i share similar sentiments.
I mean this is fascinating stuff. I've been posting in this thread for almost two years now and we are finally getting benchmarks for these two consoles. These are perhaps not the most accurate comparisons, but it's great to see stuff like this.

At the end of the day, benchmarks like this help us understand both consoles, and their deficiencies. Xbox is clearly outperforming the PS5 here and yet Remedy went with identical settings due to framedrops galore during gameplay. That indicates a bottleneck somewhere in the system, and these benchmarks can only help us figure out where that bottleneck is and if tools can fix it.

Someone on era mentioned how CPU logic is likely turned off during the photo mode and the GPU is pretty much in charge of rendering and nothing else. That could indicate an issue with the CPU, the I/O or even the ram bandwidth. Someone here, i forget who, mentioned that the lower speed ram in the XSX might bottlenecking the faster and larger pool of ram any time its accessed. Could that be the reason? because the gpu seems to be doing fine on its own.
 

JonkyDonk

Member
I mean this is fascinating stuff. I've been posting in this thread for almost two years now and we are finally getting benchmarks for these two consoles. These are perhaps not the most accurate comparisons, but it's great to see stuff like this.

At the end of the day, benchmarks like this help us understand both consoles, and their deficiencies. Xbox is clearly outperforming the PS5 here and yet Remedy went with identical settings due to framedrops galore during gameplay. That indicates a bottleneck somewhere in the system, and these benchmarks can only help us figure out where that bottleneck is and if tools can fix it.

Someone on era mentioned how CPU logic is likely turned off during the photo mode and the GPU is pretty much in charge of rendering and nothing else. That could indicate an issue with the CPU, the I/O or even the ram bandwidth. Someone here, i forget who, mentioned that the lower speed ram in the XSX might bottlenecking the faster and larger pool of ram any time its accessed. Could that be the reason? because the gpu seems to be doing fine on its own.
That's not entirely true though, CPU is involved in the BVH calculations that have to be done for RT, which I forgot earlier when considering this test. And as others mentioned, the XSX GPU is not better in every way. It has higher compute power but in other areas like fill-rate PS5 pulls ahead. Then there are the memory and I/O differences. All of these factors matter in the final performance of the system, and it's going to differ moment to moment based on specific demands of the game.

I'm also interested in whats causing those other non-GPU related stutters on Xbox, which aren't present on PS5. Could that be the I/O speed of the PS5 showing it's advantages? I wish DF would investigate these things too, instead of just being so focused on GPU heirarchy tests all the time.
 
Last edited:

Great Hair

Banned
You guys and your conspiracy theory; it would be highly illegal for MS to be paying DF for their opinions.

Microsoft > Machinima > Influencer
 

SlimySnake

Flashless at the Golden Globes
That's not entirely true though, CPU is involved in the BVH calculations that have to be done for RT, which I forgot earlier when considering this test. And as others mentioned, the XSX GPU is not better in every way. It has higher compute power but in other areas like fill-rate PS5 pulls ahead. Then there are the memory and I/O differences. All of these factors matter in the final performance of the system, and it's going to differ moment to moment based on specific demands of the game.

I'm also interested in whats causing those other non-GPU related stutters on Xbox, which aren't present on PS5. Could that be the I/O speed of the PS5 showing it's advantages? I wish DF would investigate these things too, instead of just being so focused on GPU heirarchy tests all the time.
I just saw this. Lance McDonald who is the guy who made the Bloodborne 60 fps unofficial patch, posted this comment.

EteX2DSXAAMOwbq


So according to him, the CPU load is miniscule even with RT on.
 

SlimySnake

Flashless at the Golden Globes

Microsoft > Machinima > Influencer
Microsoft also has a MVP program from their fans. They get flown out to Microsoft headquarters and party with the xbox crew. Several gaffers were Xbox MVPs. They are now over at era modding the forum. Judge is one of them.
 

JonkyDonk

Member
I just saw this. Lance McDonald who is the guy who made the Bloodborne 60 fps unofficial patch, posted this comment.

EteX2DSXAAMOwbq


So according to him, the CPU load is miniscule even with RT on.
I did see that, but he doesn't mention RT. He could have been looking at photo mode in the non-RT mode. All I know is that I've learned from DF themselves that CPU is involved quite a bit more than we might think in RT calculations. But maybe that doesn't apply to a still image? Or maybe when the CPU has no other work to do, it has plenty of headroom for RT? Why would Alex think there might be a CPU bottleneck, is he just completely wrong there?
 

JonkyDonk

Member
Saw these shots on reddit posted on era.

picture5control.png


picture3control.png


picture4control.png


picture2control.png
I don't know what to make of that. I can see the differences in these images but then the settings are supposedly identical. Alex didn't bother putting his magnifiying glass to this game because he got that email from Remedy about the settings. Maybe all the normal settings are identical but the RT settings are different?
 

IntentionalPun

Ask me about my wife's perfect butthole

Microsoft > Machinima > Influencer

And they got in trouble...
 

SlimySnake

Flashless at the Golden Globes
So, pointless comparison if he’s right about the workload?
depends on the point.

I did see that, but he doesn't mention RT. He could have been looking at photo mode in the non-RT mode. All I know is that I've learned from DF themselves that CPU is involved quite a bit more than we might think in RT calculations. But maybe that doesn't apply to a still image? Or maybe when the CPU has no other work to do, it has plenty of headroom for RT? Why would Alex think there might be a CPU bottleneck, is he just completely wrong there?
i might be able to test this on PC. i dont have an AMD card but i can run the same scene with RT on and off, cap it at 30 fps and see if the CPU usage increases in the RT mode.
 

JonkyDonk

Member
depends on the point.


i might be able to test this on PC. i dont have an AMD card but i can run the same scene with RT on and off, cap it at 30 fps and see if the CPU usage increases in the RT mode.
Actually someone in the other thead about this benchmark just posted a test showing almost the same CPU usage in game and in photo mode. If that's true, then it's very interesting.
 

SlimySnake

Flashless at the Golden Globes
Guys i'm so in love with this newly made wallpaper (made by CD Art Director)

Ultra excited about Nexgen Lara Croft Tomb Raider, CANNOT Wait!!!

EtPgzsGVcAAq9nb
i am still shocked they never gave her dual pistols. Especially after how they teased it in the ending QTE of the first game. I actually liked the reboot and Rise, but the series was a massive missed opportunity. The series can be grounded and still have dual wielding, fast paced run and gun shooting and vehicles. I mean thats literally what Uncharted does.
 

LucidFlux

Member
Saw these shots on reddit posted on era.

picture5control.png


picture3control.png


picture4control.png


picture2control.png

Curiouser and curiouser.

You can clearly see the missing hair and difference in the fog/lighting. I'm not entirely sure I'm seeing anything in those other shots though.

I wonder... while the game settings are confirmed the same but what if for some reason photo mode has different settings enabled on the two machines? There is already this bug unlocking frame rate so it wouldn't surprise me.
 

SSfox

Member
Forgive my ignorance, but is a next-gen Tomb Raider in the works?
No no if it was you know that the Internet would have blown lol They confirmed they started working on the next TR game, and this is a new Artwork made by the Art Director of the game but it's not necessarily relate to the game, still maybe it hint that it will go back to the root of the original TR games by Core. Would make sens since they planning to use the TV show to explain how Lara transitioned from young reboot lara to more mature experienced OG Lara.
 

Lysandros

Member
Curiouser and curiouser.

You can clearly see the missing hair and difference in the fog/lighting. I'm not entirely sure I'm seeing anything in those other shots though.

I wonder... while the game settings are confirmed the same but what if for some reason photo mode has different settings enabled on the two machines? There is already this bug unlocking frame rate so it wouldn't surprise me.
Third picture's volumetric lighting difference is also very obvious.
 
I just saw this. Lance McDonald who is the guy who made the Bloodborne 60 fps unofficial patch, posted this comment.

EteX2DSXAAMOwbq


So according to him, the CPU load is miniscule even with RT on.
He's right. Only GPU stress, essentially same data on caches. CPU in fixed workload.

In Gameplay you have streaming into RAM, caches, CPU and GPU varying workloads.

PS5 has scrubbers, cohency, maybe unified l3 cache. In gameplay it's more balanced.
 

SSfox

Member
i am still shocked they never gave her dual pistols. Especially after how they teased it in the ending QTE of the first game. I actually liked the reboot and Rise, but the series was a massive missed opportunity. The series can be grounded and still have dual wielding, fast paced run and gun shooting and vehicles. I mean thats literally what Uncharted does.
Maybe they didn't find a way to make it appealing with Reboot gameplay, while saving this for the next trilogie, in any case it didn't bother much as much as i'm big Core Design TR. TR reboot was their way to make the game like they want all the way, in a sens imo it's good they give OG lara a break with her dual pistols, short shorts, T-rex and all what made OG Lara what she was. It's like when James Bond they stopped making him using gadgets in few movies, it's pissing at first for the fans but then you think it's not that bad they took a little break on those iconic things he's been using for so many movies and so long -and so same with OG Lara- a little moment, so when they brought those back it will make you go WOW OMG with the excitements and all. That's how i see it lol
 
Last edited:

Lysandros

Member
Yep I mentioned that one. It's images 2 and 4 I'm not noticing much, if any difference in, but regardless there is clearly something funky going on.
RT reflections seem indeed more noisy to me in the second image on XSX, i don't think it's placebo. But for the last image i agree, i can't differentiate.
 

ethomaz

Banned
He's right. Only GPU stress, essentially same data on caches. CPU in fixed workload.

In Gameplay you have streaming into RAM, caches, CPU and GPU varying workloads.

PS5 has scrubbers, cohency, maybe unified l3 cache. In gameplay it's more balanced.
You probably didn't saw usual Xbox fans asking for screen shot lol
Or trying to say say he is not in the Control credits lol

I'm not sure even how to reply to them... seems like we are losing half of the brain.
 

HoofHearted

Member
It’s basic economics, if you don’t understand how supply and demand determines prices... you’re either too young to have learned or you skipped school and it shows. Being ignorant isn’t cool man.

By the way, as far as UK goes, for 2020, all Xbox consoles were behind Switch, PS5, PS4. In that order.

I'm fully aware of how basic economics works. I'm also fairly confident that the last indicator that economists would look to for "demand" would be eBay. Hell - let's include Facebook Marketplace, OfferUp, or whatever other latest hip app is used to offload used stuff...

I honestly could care less about which console sold more than the other - they're both selling as fast as they're made. However, you clearly do care for some odd reason???

The fact is - ALL of the consoles are currently in active demand. Sony, MS, and Nintendo can't make them fast enough to meet the supply.

We won't know actual demand until there's an overabundance of supply.

I find it rather silly that someone posts a tweet about how Xbox sold more than PS for a period of time, and immediately everyone jumps in to defend Sony because of... eBay pricing???

My point is - prices on eBay fluctuate like farts in the wind. Prices also fluctuate daily, or week to week. Primarily because some people are too "ignorant" and pay more than necessary...

Grow up and stop championing a plastic box.

Here's a hint - neither Sony or MS execs really care about the console "war" - it's not on their radar - they just want you to buy their shit...
 
Status
Not open for further replies.
Top Bottom