sncvsrtoip
Member
but without bug and with dynamic 4k kicking in ps5 is close to 60fps in this scene
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:Call of Duty Black Ops Cold War - PS5 vs PC - Settings And Performance Analysis
- Thread starter SmokSmog
- Start date
ethomaz
Banned
Because I believe the bug exists in that scene.
He needed the bug to force 4k native for comparison.
Last edited:
Rikkori
Member
Must've missed hearing that, but makes sense.
ethomaz
Banned
The game actually runs that scene at dynamic 1440-2160p (not exactly 1440p but similar in pixels 1920x1800) holding solid 60fps.
But due a bug on PS5 code it sometimes runs with Dynamic res disabled... so 2160p native with fps around 45.
That is how he could make the comparison in the video.
If you reload the checkpoint the bug is gone and DRS enabled again.
Last edited:
rodrigolfp
Haptic Gamepads 4 Life
Impossible for a 2021 gaming machine miss basic features like that that even 90s computers have. oh wait....
Fredrik
Member
Yeah dynamic resolution is messing up most console vs PC comparisons, settings differences is another thing but DF and Alex did good here and PS5 owners should be impressed by what their $400-500 box is giving them.
I nearly upgraded my GPU to a 3070 recently because 3080 is impossible to find. Would’ve cost me about $800, and would’ve pissed me off after seeing this comparison. Going by this game it seems like DLSS is the only thing that card has that pulls it significantly ahead of a PS5.
mckmas8808
Mckmaster uses MasterCard to buy Slave drives
Good points.
PaintTinJr
Member
Does a DLSS solution render the depth buffer (zbuffer) for picking (collision detection with the crosshair) at the native resolution or the lower resolution?
If it is the lower resolution then it isn't pulling ahead at all - and DLSS can then be categorized as inferior for an FPS game, versus native - because picking accuracy is in essence proportional to zbuffer and rendering resolution. You can't accurately reconstruct a picking point that wasn't there - unless you render polygons with polygonID colours and then do some over-estimated hit test in model space.
Fredrik
Member
It’s a developer issue and it’s dumb. Console games with a PC settings menu would solve a lot. Changing the FOV can help against motion sickness, graphics settings can help against stutter from not reaching the framerate target. Instead of minor tweaks people go to twitter to scream at the devs so they’ll patch the games, then the patch break something else, more screaming, another patch, another scream, etc
SlimySnake
Flashless at the Golden Globes
Gotta give it to Alex. This mother fucker has an insane eye for detail and has managed to come up with some of the most detailed comparisons I have ever seen. Dont like the guy's personality but there is no denying he is really good at his job.
Respect.
Respect.
SlimySnake
Flashless at the Golden Globes
Pretty sure Battlefield 1 and BFV had FOV sliders. I know Destiny added it recently and the Borderlands PS5 patch also allows it.
This isnt a machine specific feature.
rodrigolfp
Haptic Gamepads 4 Life
No one said it is.
SlimySnake
Flashless at the Golden Globes
PaintTinJr
Member
I'll have to put the disc back in later tonight, as I was sure it had a FOV slider when I turned off all the aim assist for the game and switch my controls - before playing through the entire campaign and really liking it, and being graphically impressed. but being a last-gen game that isn't on the disc (180GB install from disc and online later), it could have been just the menus in the multiplayer I was able to adjust that.It’s a developer issue and it’s dumb. Console games with a PC settings menu would solve a lot. Changing the FOV can help against motion sickness, graphics settings can help against stutter from not reaching the framerate target. Instead of minor tweaks people go to twitter to scream at the devs so they’ll patch the games, then the patch break something else, more screaming, another patch, another scream, etc
Fredrik
Member
I don’t have a DLSS capable GPU so no first hand experiences here. From what I’ve heard DLSS 1.0 had some issues, wonky image artifacts, but DLSS 2.0 is nuts and things can actually look better with DLSS ON and 1440p native resolution than DLSS OFF and 4K native resolution.
I have no clue how it’s done.
The drawback is that it’s not added to all games.
rodrigolfp
Haptic Gamepads 4 Life
Where did I say it's a "specific" machine feature? This is not a hardware feature. Sorry if it "sounded" like that.
Last edited:
Md Ray
Member
But it's not doubling the bandwidth.
Fredrik
Member
No idea I was just ranting about my annoyance that devs are removing settings menus on console, it’s not a console issue, it’s a console developer issue, and I hate it.
ethomaz
Banned
I believe he tried to match PS5.
Games doesn’t use 8 cores in consoles.
Last edited:
silent head
Member
1920x1800 - 2160P
Both PS5 and Series X in the ray tracing mode use a dynamic resolution with the highest resolution found being 3840x2160 and the lowest resolution found being 1920x1800
ethomaz
Banned
So it drops more horizontally than vertically... not proportionally.
1920x1800 is similar to 1440p in pixels being rendered... I will fix my comment... thanks.
Last edited:
CPU might be the limiting factor here for 3070, just saying..... unless they do further test, no one knows
kyliethicc
Member
So..
"An RTX 2070 Super can't match PlayStation 5 - in fact, it's 20 per cent slower. On the AMD side of things, I found the RX 6800 XT's result to be off-pace - it has 72 compute units vs the 36 inside PlayStation 5, it's based on the same architecture, and clock speeds are broadly equivalent, yet it delivered just 29.4 per cent of extra performance."
So the PS5 is at least around the 2080 or 3060 Ti in rasterized gaming. Those GPUs cost as much or more than the PS5. And the 6800 XT is more powerful (of course), but that 1 component costs like $700, while the entire PS5 system can be had for $400.
Pretty good job by Cerny and da boys.
"An RTX 2070 Super can't match PlayStation 5 - in fact, it's 20 per cent slower. On the AMD side of things, I found the RX 6800 XT's result to be off-pace - it has 72 compute units vs the 36 inside PlayStation 5, it's based on the same architecture, and clock speeds are broadly equivalent, yet it delivered just 29.4 per cent of extra performance."
So the PS5 is at least around the 2080 or 3060 Ti in rasterized gaming. Those GPUs cost as much or more than the PS5. And the 6800 XT is more powerful (of course), but that 1 component costs like $700, while the entire PS5 system can be had for $400.
Pretty good job by Cerny and da boys.
Md Ray
Member
Given that bandwidth is also shared between the GPU and CPU on PS5 (and other consoles) as compared to a PC with dedicated GPU with its own dedicated bandwidth, these results are fascinating. Even when you take PC's higher settings into account, PS5 is likely performing like a 3060 Ti/RTX 2080S, more or less.
rodrigolfp
Haptic Gamepads 4 Life
more like pretty bad Activision job on pc. It's like this since CoD Ghosts.
Last edited:
kyliethicc
Member
well with the cache it has, it sorta does have 2x bandwidth.
kyliethicc
Member
irony is they're Nvidia sponsored titles
Md Ray
Member
CPU limit? They (Alex) generally uses 10900K for these kinds of tests.
SlimySnake
Flashless at the Golden Globes
not for this one. he used the 3060 for these tests because he thinks the game is CPU bound with ray tracing on.
if you want to compare the PS5 GPU to the PC GPUs, you have to use the same CPU.
rodrigolfp
Haptic Gamepads 4 Life
I would ask Nvidia for some of my money back.
Md Ray
Member
Other scenes were probably locked 60fps and working as intended. Those frame-rate dips to mid-40s on PS5 occur due to a bug that disables dynamic resolution and thus runs the game at native 4K.
S0ULZB0URNE
Member
2080S
phat_crazy
Member
Considering what Sony's first part devs had to work with last gen, we are going to see unbelievable gameplay sequences and set pieces
Studios like ND will make the PS5 punch well above its weight
So if the PS5 is basically a 2070S or 3070, that is incredible considering it is a closed system that will be maximized by Sony's talented developers
Last edited:
kyliethicc
Member
There will not be a PS5 Pro until Sony is ready to cut the price of the current PS5. That won't be until 2023 or 2024, if ever. Sony is having a hard enough time making enough PS5s and getting them to market. Don't expect any new hardware from them for years.
If you really need more "power" than the PS5, which can run games at 4K@60 Hz, then try PC gaming. Its fun to discuss all the power & specs but they're not really the point of consoles. Its about affordable gaming for a mass audience. 1 simple box to buy for 6-8 years of fun.
Md Ray
Member
Give them all the settings to tweak and they'll still scream like the dimwits on PC whine and cry "UNOPTIMIZATION" after cranking up all the settings on their low-mid range PC.It’s a developer issue and it’s dumb. Console games with a PC settings menu would solve a lot. Changing the FOV can help against motion sickness, graphics settings can help against stutter from not reaching the framerate target. Instead of minor tweaks people go to twitter to scream at the devs so they’ll patch the games, then the patch break something else, more screaming, another patch, another scream, etc
Rudius
Member
The 3090 is a bit over 80% faster than the PS5 with that same CPU, so it's apparently sufficient. Also, this shows that the 36 teraflops don't mean much in real life performance. It not even doubles the PS5 performance.
Spukc
always chasing the next thrill
You must be new to pc gaming
SlimySnake
Flashless at the Golden Globes
Pretty impressive performance by the PS5.
A 3070 is offering only 8.6% more performance despite being 20.3 tflops or roughly 100% more tflops compared to the PS5.
A 3090 is offering 81% more performance despite being 36 tflops or roughly 260% more tflops compared to the PS5.
Comparing it to and AMD GPU, the 20 tflops 6800xt is offering only 29% more performance for 100% more tflops.
This chart here shows the relative performance of a 3070. It offers 11% more performance than a 2080 Super and 10% more performance than the 3060 Ti. That means the PS5 is performing a little better or on par with the 2080 Super and the 3060 Ti. At least for this game. Pretty fucking impressive.
A 3070 is offering only 8.6% more performance despite being 20.3 tflops or roughly 100% more tflops compared to the PS5.
A 3090 is offering 81% more performance despite being 36 tflops or roughly 260% more tflops compared to the PS5.
Comparing it to and AMD GPU, the 20 tflops 6800xt is offering only 29% more performance for 100% more tflops.
This chart here shows the relative performance of a 3070. It offers 11% more performance than a 2080 Super and 10% more performance than the 3060 Ti. That means the PS5 is performing a little better or on par with the 2080 Super and the 3060 Ti. At least for this game. Pretty fucking impressive.
Md Ray
Member
I just spoke to him. He did use 10900K to be GPU bound with the 3070. Same for the other GPUs he tested: 3090, 2080 Ti, 6800 XT, etc.
Last edited:
Gold_Loot
Member
yes, BO Cold War has a FOV slider on PS5.
Fredrik
Member
They can still have standard console settings as default settings and just put the other settings in an advanced settings menu which those who don’t know what they’re doing won’t dare touching.
Instead of going back and forth with patches removing this and that and getting users annoyed it would be better to let the user decide, maybe they don’t actually care if they remove some details nobody will ever notice without a paused 400x zoomed DF video.
rodrigolfp
Haptic Gamepads 4 Life
Why are you comparing flops between AMD and Nvidia?Pretty impressive performance by the PS5.
A 3070 is offering only 8.6% more performance despite being 20.3 tflops or roughly 100% more tflops compared to the PS5.
A 3090 is offering 81% more performance despite being 36 tflops or roughly 260% more tflops compared to the PS5.
Comparing it to and AMD GPU, the 20 tflops 6800xt is offering only 29% more performance for 100% more tflops.
This chart here shows the relative performance of a 3070. It offers 11% more performance than a 2080 Super and 10% more performance than the 3060 Ti. That means the PS5 is performing a little better or on par with the 2080 Super and the 3060 Ti. At least for this game. Pretty fucking impressive.
Plus. This game is like the Crysis situation. Does not matter the power of the pc, it will not use the power right if not coded properly.
They think things like SotC and Last Guardian with their shitty performance is wonderful, so no.
Last edited: