Support NeoGAF

[Leyasu]

Slow and Wide vs Fast and Narrow, very embarrassing for DF, benchmarking pause menu's vs actually gamplay

Every analysis done by every site just takes a few samples and posts their results. As we saw with this game both consoles drop frames. What it is like for the whole games and percentages or swings for either console is hard to know. I can't imagine any site or youtuber playing the whole game on consoles, then rewatching it to see the performance deltas. I don't know. It didn't seem to me that they were reaching.

They clearly explain what they are doing and why it is for them a good way to benchmark the consoles performance.

 

[Thirty7ven]

Why am I not surprised that people are showing their ignorance on the fact that photo mode performance doesn’t translate 1:1 to moving gameplay.

Embarrassing. This is the problem with DF, they actually don’t know how engines work.

 

[MilkyJoe]

We need good kart games on the next gen twins

I'm not going to say no

Kart games rule, racing games rule,

 

[Krappadizzle]

This is it, guys. I am buying an Xbox Series X for playing Control.

After all, you do need very high frame rates when taking photos in Photo Mode

 

[mansoor1980]

photo mode only uses the faster portion of the xsx ram pool

 

[sircaw]

They finally found a way to fix the stutter i see.

 

[Leyasu]

I don't know or give a shit what people said last gen, or who 'you guys' is. I wasn't here and never spoke about such things.

All I'm seeing is what's happening right now: Xbox fanboys spent the last year talking up 'massive' power differences between these systems, and they were told by more reasonable people that it's only an 18% GPU difference, and now here we have a GPU test showing an average of 16% difference, and Xbox fanboys are acting like this is some huge revelation and victory.

Yet every launch game port that has performed better on the PS5 has had the same morons talking about bottlenecks and other stuff that they have no clue about, or repeating Cernys speak cache scrubbers (when they don't even know their real world benefit or can give anything to quantify their advatange, whilst simultaneously telling people that they shouldn't have listend to MS's marketing.

Two wrongs don't make a right

 

[onnextflix5]

Dose this not prove there is bottleneck somewhere on the xsx as the extra power can not used during gameplay due to this bottleneck?

 

[Mister Wolf]

So a benchmark that confirms everything that we knew about the GPU differences?

Remedy should give the XSX a full 60 mode its hovering close enough a relaxed DSR could get this game that high.

This generation/forum is going to get really messy as more games take advantage of the extra power available to the XSX.
Im probably gonna have to take self imposed ban whenever a big multiplatform title shows up.

I don't think their Northlight Engine supports DRS. Certainly doesn't on PC.

 

[Sw0pDiller]

great video! Finally a really great advantage for series x to show it's theoretical advantages. Alas only in theoretical setting and not gameplay. May be a sign of things to come. Maybe series x might get 60fps quality modes over a 30 fps mode on ps5. Weird that in most games and thus normal circumstances ps5 seems to pull ahead.

 

[Rea]

Rendering static scenes which basically are duplicate frames and called them benchmark... that's so embarrassing, the real stress test, AKA. corridor of doom, both perform the same, so what's the story?. DF is making themselves more embarrassing day by day, also Richard commented on PS5 chip being smaller, did We have ps5 dieshot yet? No? How does he know, guesswork.

 

[TGO]

Photo mode? umm okay.
And yet

 

[Topher]

Rendering static scenes which basically are duplicate frames and called them benchmark... that's so embarrassing, the real stress test, AKA. corridor of doom, both perform the same, so what's the story?. DF is making themselves more embarrassing day by day, also Richard commented on PS5 chip being smaller, did We have ps5 dieshot yet? No? How does he know, guesswork.

He was referring to number of CUs.

 

[MadchesterManc]

Never expected a photo mode to be such a thirst quencher.

 

[Mister Wolf]

What i find interesting is that A you would need a 2070s or so which is more inline with expectations and B the 16 percent diff seems about right with the paper specs.

The PC cant do console settings. They are too low and custom. One example is the raytraced reflections running at twice the resolution of consoles with no way to lower them. Think about how much performance the console's are saving dropping them by half.

 

[scydrex]

Damnnnn Remedy got the tools in photo mode for the XSX!

 

[x@3f*oo_e!]

Does require a minimum 40hz though... at least with my TV and every other one I've seen. If the minimum was 40 fps I completely agree this should be an option.

Most are sub 20Hz

Our Monitor Motion Tests: Refresh Rate

A monitor’s refresh rate is the frequency at which it can display frames every second. It's measured in Hz, and a higher refresh rate produces smoother and clearer motion if properly used.

www.rtings.com

 

[JonkyDonk]

Yet every launch game port that has performed better on the PS5 has had the same morons talking about bottlenecks and other stuff that they have no clue about, or repeating Cernys speak cache scrubbers (when they don't even know their real world benefit or can give anything to quantify their advatange, whilst simultaneously telling people that they shouldn't have listend to MS's marketing.

Two wrongs don't make a right

I don't understand what your point is. Those games are performing better on PS5 and there must be a reason for it, so people speculate about those reasons. That doesn't change the raw specs, they are what they are. The point is that a lot more goes into console performance than just raw theoretical compute power, this is what people have been trying to tell Xbox fanboys all along and they don't get it. All this test shows, is that the GPU is indeed bigger as we've always known, but it's in abstracted conditions of a photo mode which is very different from actual gameplay. As Richard said in the video, it's academic.

 

[mcjmetroid]

My god. Will they EVER stop going on about this game...

 

[Antony1992]

So there are no circumstances where Ps5 has some advantages over SX this time, like in Hitman 3 (it was just one small area) ? So you are telling me that now Ps5 fanboys can only accept defeat without excuses?
Oh, damn.

 

[Alebrije]

Will wait for the RE8 video...
But no matter which console has better "performance" the difference on this generation is mininal, basically you have the same gameplay experience on both.

Its different from PS4 vs Original Xbox where the gap was bigger.

I do not remember a DF video where Photo mode was the benchmark for analysis...but maybe they did this before several times and I am wrong....

Control plays basically the same in both consoles..you are not loosing anything no matter on which console you play it

 

[ethomaz]

Will wait for the RE8 video...
But no matter which console has better "performance" the difference on this generation is mininal, basically you have the same gameplay experience on both.

Its different from PS4 vs Original Xbox where the gap was bigger.

I do not remember a DF video where Photo mode was the benchmark for analysis...but maybe they did this before several times and I am wrong....

Control plays basically the same in both consoles..you are not loosing anything no matter on which you play it

He will call it is a Japanese engine.

This video is actually interesting because it shows there are some bottleneck (probably CPU?) in Series X that doesn’t allow to use the GPU power.

We are speculation that in others games without evidence... this video is the first real evidence of Series X being bottlenecked.

PS. Thinking a bit about that can be the reason that even having a stronger GPU the power draw is similar in gaming.

 

[Black_Stride]

Rendering static scenes which basically are duplicate frames and called them benchmark... that's so embarrassing, the real stress test, AKA. corridor of doom, both perform the same, so what's the story?. DF is making themselves more embarrassing day by day, also Richard commented on PS5 chip being smaller, did We have ps5 dieshot yet? No? How does he know, guesswork.

You obviously dont know how framerates work? Do you assume the game stops rendering stuff when the camera is static?
So the starting grid of a racing game is running at 0fps?

DFs tools can measure per pixel differences - games that use TAA have minute differences in each new frame even if a camera is still and the image is perceptually very similar to the last.

Your framerate doesnt drop to zero just because the camera is static.
This is exactly why softwares like TRDROP are heavily discouraged in the analysis community, its too dumb to know the difference between 0fps and a static image being rerendered multiple times.

Photo mode? umm okay.
And yet

It drop 2 frames?
OMG stop the presses

 

[Mr Moose]

So there are no circumstances where Ps5 has some advantages over SX this time, like in Hitman 3 (it was just one small area) ? So you are telling me that now Ps5 fanboys can only accept defeat without excuses?
Oh, damn.

Only during gameplay.

 

[TGO]

So going forward are we gonna see Photo Mode benchmarks?

 

[Interfectum]

These games are so close to each other literally no one here would know the difference unless DF told them. Sony fans are taking it pretty hard in here though... ya'll should start talking about kraken and noodles again.

 

[Elias]

and this just the raw advantage of the gpu without the use of vrs, maching learning dedicated silicon, and sfs.

 

[Bernd Lauert]

Just here for the salt.

Looks like the cache scrubbers didn't pull through this time. Needs more Cerny sauce.

 

[FireFly]

I'm just saying if the AMD RX 6800 (with 60 CUs) got benchmarked against the upcoming RX 6700 XT (with 40 CUs), the 6800 is gonna win in RT benchmarks with all other settings equal. Same with comparing Nvidia cards. More RT cores = more FPS in RT.

So the XSX can average like 5 FPS higher Control RT in PHOTO MODE lol ... but PS5 has what counts. The games. (And its $100 cheaper.)

Sure. My point is just that the delta is not as big as the absolute difference in CUs, because the PS5 has up to a 23% clock speed advantage.

 

[asustitan]

Apples for Apples in this one

 

[SplunkyMunkey]

Of course ps5 will do well on 60fps even 120fps considering the cpu on the ps5 is unified and of course series x the fps will drop in scenes where there is lots going on as the gpu on xsx is slower compared to ps5. As every1 saying the 52cu will do well in raytraced games. And as for hitman 3 ps5 could easily do full 4k but i think its the glacier engine that is crap

 

[martino]

So a benchmark that confirms everything that we knew about the GPU differences?

Remedy should give the XSX a full 60 mode its hovering close enough a relaxed DSR could get this game that high.

This generation/forum is going to get really messy as more games take advantage of the extra power available to the XSX.
Im probably gonna have to take self imposed ban whenever a big multiplatform title shows up.

would it helps for scenes like/near corridor of doom and how many of them there is ?

 

[OverHeat]

Consoles Warriors are so easy to spot around here its crazy...Cant believe that on a gaming forum where the most hardcore gamers are posting they still only play on one platform. Screw Fanboys

 

[Portugeezer]

Photo mode is not a perfect representation, but this is the sort of difference we would have expected before the consoles launched, we just haven't seen much of it in actual reality.

 

[Rikkori]

I like how they pretend not to know this RT implementation was basically done by a couple of Nvidia engineers who worked side-by-side with Remedy without whom this game would have no RT. But hey, gotta blame AMD somehow, it's that missing pesky magical driver who is going to re-write the game code that's the issue.

 

[x@3f*oo_e!]

DF is making themselves more embarrassing day by day, also Richard commented on PS5 chip being smaller, did We have ps5 dieshot yet? No? How does he know, guesswork.

308mm2 btw

 

[Mister Wolf]

You will only see that GPU advantage in multiplats running at high resolutions locked and not hiding behind dynamic resolution scaling. And when I say "high" I mean 1800p and above. You may see it in RE8 which is targeting 60 while running at 4K(not dynamic) and using raytracing.

 

[ChuckeRearmed]

How did sony manage to produce such brainwashed and religiously loyal fans ? The loyalty for this corporation is on par with religious fanatics

It is similar to Apple in that regard. Basically Sony was successful over several generations creating a bunch of zealous people.

Interesting thing is that Nintendo did not have these groups - probably because it's main community either younger or older.

 

[skneogaf]

This game could definitely be improved with added modes like ray tracing and unlocked fps as they can guarantee to stay within vrr range.

 

[Snake29]

So a photo (mode) finish for the XSX,....got it...

 

[Heisenberg007]

Only during gameplay.

But that's not important. Nobody buys a game to play it.

 

[Leyasu]

I don't understand what your point is. Those games are performing better on PS5 and there must be a reason for it, so people speculate about those reasons. That doesn't change the raw specs, they are what they are. The point is that a lot more goes into console performance than just raw theoretical compute power, this is what people have been trying to tell Xbox fanboys all along and they don't get it. All this test shows, is that the GPU is indeed bigger as we've always known, but it's in abstracted conditions of a photo mode which is very different from actual gameplay. As Richard said in the video, it's academic.

My point was that people using these launch games as a basis for anything is wrong.

Still, I agree with what you are saying, and think that some games will perform better on one console and others better on the another for the whole gen.

 

[ethomaz]

It is similar to Apple in that regard. Basically Sony was successful over several generations creating a bunch of zealous people.

Interesting thing is that Nintendo did not have these groups - probably because it's main community either younger or older.

Nintendo fans are the most loyal... just look at the GAF threads.

 

[fatmarco]

Do we know the OS cost of these systems? Specifically for the CPU in relation to Sx. Could things like quick resume etc. be an issue? Do we know how expensive that is?

 

[AnotherOne]

I hope you're happy Riky.

 

[Heisenberg007]

This game could definitely be improved with added modes like ray tracing and unlocked fps as they can guarantee to stay within vrr range.

Games can't be designed with VRR in mind. You can't expect devs to cater to roughly 2% of gamers and give the rest of the 98% an unstable game.

 

[ChuckeRearmed]

Nintendo fans are the most loyal... just look at the GAF threads.

But they don't have Defense Force units. They live in the own world basically.

 

[Rea]

You obviously dont know how framerates work? Do you assume the game stops rendering stuff when the camera is static?
So the starting grid of a racing game is running at 0fps?

DFs tools can measure per pixel differences - games that use TAA have minute differences in each new frame even if a camera is still and the image is perceptually very similar to the last.

Your framerate doesnt drop to zero just because the camera is static.
This is exactly why softwares like TRDROP are heavily discouraged in the analysis community, its too dumb to know the difference between 0fps and a static image being rerendered multiple times.


It drop 2 frames?
OMG stop the presses

What are you talking about, I'm not disagreeing with DF for frame rates, but that does not change the fact that rendering the same image over and over again is not more taxing than rendering different frames every seconds. Also the photo mode is never a representative of the actual gameplay. That's just a bad way of benchmarking.

 

[Spokker]

This is a good test because it probably takes the CPU, and more importantly, any I/O issues out of the equation. It lends credence to the idea that there are other issues preventing the Series X from reaching its full potential. The console has been a letdown based on the way Microsoft hyped it up as the system of "no compromises," and a win is a win.

 

[spyshagg]

There are two victories in this bench, but one is not apparent without thinking what is happening there.