• 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.

Bo_Hazem

Banned


Was this posted, thread worthy..


 

Bo_Hazem

Banned
I did. Not sure what you seen to say such things.

fQukxBC.jpg
 

assurdum

Banned
Yes. And there are moments which even series X outperformed ps5 in this mode. In the 60 FPS mode during the gameplay it stay literally above the ps5 almost the entirely time. Now we are talking of ridiculous difference but still shouldn't be ignored. Call the ps5 version of DMC5 better it's definitely a stretch. Of course we aren't not talking of statics so I take it as a grain of salt.
 
Last edited:

cragarmi

Member
Even if you consider the 120 mode on XSX bugged, the PS5 is clearly more performant across the board, there is more than just 'tools' holding the Xbox back here.
 

Krisprolls

Banned
What yous are explaining makes perfect sense if you are trying to compare PC GPU vs PC GPU. However throwing a console into the mix kinda disrupts the whole comparison doesnt it? Because essentially you are no longer comparing GPU vs GPU but system vs system? Also the resolution isnt even the same, this whole comparison is kinda dumb if you ask me. 🤷‍♂️

Alex is a PCMR and Xbox fanboy, you won't ever get any fair comparison from him. DF shouldn't be trusted as much as they are. They conveniently forget the points which don't fit the agenda.

There's a game (Valhalla I think) where even PC version stutters in cutscenes like XSX does. Only PS5 version is smooth.

Basically PS5 is the best version even compared to any PC.

PS5 has hardware optimizations and features that no PC has simply because it's a dedicated game machine and not a standard PC. It can't be solved by adding more cpu or gpu power. It's just making the pc more bottlenecked elsewhere.
 
Last edited:
How this test is invalid? He wanted to compare PS5 GPU with PC counterparts, so he make sure that he is not going to be bottlenecked by CPU.
But how does he know this? The only way to know is to do a CPU comparison in the same scene. Otherwise it's pure conjecture. We are talking about a mobile CPU at 3.5ghz VS one of the strongest CPU in the world at 5ghz.
 
Last edited:

Bo_Hazem

Banned
Even if you consider the 120 mode on XSX bugged, the PS5 is clearly more performant across the board, there is more than just 'tools' holding the Xbox back here.

Just wanna tell you, my DS controller never died on me, not even close with gaming sessions up to 12 hours. But Charging station is wonderful if you finish or have a small break just put it on it. Same with Pulse 3D (I charge it with a cellphone charger), the battery seems good enough on both.
 
Last edited:

Garani

Member
$499
everything above that price would have been a desaster... It was the maximal price possible, nearly no matter how fast the device is.
Sony wants to sell 100+ mio PS5, and you can do that only if you price the device on a level where normal people are buying them, not just some nerds like us.

And I still think that $499 is still too high. But for the moment people are buying, so that's ok.

The name ... Dirt 5 - The Next Generation Technical Analysis - PS5 | Series X | Series S.

Maybe best to actually watch the videos then post.

I did watch the video, but not THAT one. It was late at night and I was sure to have watched it. But I see this morning that the video is not dated 6th of december, while I was sure that I watched the one dated 5th and still on processing earlier on.

Oh well, it is what it is, and I did manage to watch this version. Basically PS5 keeps a more decent geometrybut dips lower in the 120 frames, while XSX is just a pain, but it hits 120FPS, hey! :D
 

raul3d

Member
except the PS5 CPU is weaker than the Intel 10900K, so we don't know if the PS5 would perform better with a stronger CPU.

No one would benchmark 4 different GPUs using a 10900K for 3 of them and then a 3700X for the last one. It would be an invalid comparison for all 4 cards.

To benchmark against the PS5, a PC spec I'd use is:

Ryzen 3700X CPU @ 3.6 GHz
X570 motherboard, 16 GB DDR4
Nvidia RTX 3060 Ti FE or RTX 2080 FE
1 TB Samsung 980 Pro NVMe SSD
The benchmark shows the 10900K's cores utilization. I would say that you can ignore all the virtual cores, since half the cores are basically idle. The other cores are between 20-40% utilization and this is on a PC with more overhead. The PS5's CPU should have more than enough cores and frequency to show similar performance here, so not bottleneck it's GPU in this scene.

Though I agree that either testing outright with an 3700X or at least validating the tests with an 3700X to see if the performance would change, would make more sense. Even if I think that it would not change anything meaningful in this scene.

What yous are explaining makes perfect sense if you are trying to compare PC GPU vs PC GPU. However throwing a console into the mix kinda disrupts the whole comparison doesnt it? Because essentially you are no longer comparing GPU vs GPU but system vs system? Also the resolution isnt even the same, this whole comparison is kinda dumb if you ask me. 🤷‍♂️
Sure, you compare system against system; basically you always compare system against system. That's why you try to isolate the component that you want to benchmark by using a scenario where this component appears to be the bottleneck and powering up all other components to further reduce the pontential bottlenecks.

With the goal of comparing the PS5's GPU against the PC, what would you change to make the benchmark more accurate?
 

roops67

Member
Yes. And there are moments which even series X outperformed ps5 in this mode. In the 60 FPS mode during the gameplay it stay literally above the ps5 almost the entirely time. Now we are talking of ridiculous difference but still shouldn't be ignored. Call the ps5 version of DMC5 better it's definitely a stretch. Of course we aren't not talking of statics so I take it as a grain of salt.
Going by the recent Dirt5 analysis. We need to compare more than the FPS at the time , need to check other differences on screen like average-resolution geometry asset-quality shadows anisotropic-filtering ambient-occlusion and much more etc. So it's not just apples to apples comparisons anymore for the same game on both consoles
 
Last edited:

azertydu91

Hard to Kill
The benchmark shows the 10900K's cores utilization. I would say that you can ignore all the virtual cores, since half the cores are basically idle. The other cores are between 20-40% utilization and this is on a PC with more overhead. The PS5's CPU should have more than enough cores and frequency to show similar performance here, so not bottleneck it's GPU in this scene.

Though I agree that either testing outright with an 3700X or at least validating the tests with an 3700X to see if the performance would change, would make more sense. Even if I think that it would not change anything meaningful in this scene.


Sure, you compare system against system; basically you always compare system against system. That's why you try to isolate the component that you want to benchmark by using a scenario where this component appears to be the bottleneck and powering up all other components to further reduce the pontential bottlenecks.

With the goal of comparing the PS5's GPU against the PC, what would you change to make the benchmark more accurate?
Sure it show it is not fully used, but in a scientific comparisons differences are called factors and to get the best possible comparison you need to change the least possible amount of factors between comparisons (prefereably only 1), in this case interactions between CPU and GPU are another factor.
Even worse there's a lot of other factor changing like OS, RAM etc ....
 

assurdum

Banned
Going by the recent Dirt5 analysis. We need to compare more than the FPS at the time , need to check other differences on screen like average resolution and asset quality and much more etc. So it's not just apples to apples comparisons anymore for the same game on both consoles
That's why I said based on what I see in the DF video.
 
Last edited:

Bo_Hazem

Banned
Some dual sense battery feedback.

In a very unscientific, but more accurate use case, yesterday I played demon's souls using the dual sense for 11 hours non stop. By then the controller was on its last bar but still didn't give me the "battery low" message. Unfortunately I had to stop at that point, but I suspect it was probably an hour at most away from needing to charge.

Speaking of charging, using the PS5 rest mode it took exactly 1 hour 54 mins to fully charge the controller from the above position.

So at least for my case, the controller has a "11-12 hour battery life" and a "2 hour recharge time".

Makes sense as to why Sony allows you to set the usb power output for 3 hours after going into rest mode. It's basically enough to charge the controller from dead.

I have the charging station next to bed, when finished just throw it there. Buying a second controller is actually a waste as I've played for ~12 hours yet no low battery notification, same with Pulse 3D that I charge it via a cellphone charger. It sounds the same turned off and wired to DS anyway, very impressive for music. On Watch Dogs Legion the sounds are pure porn when you are in car with radio or inside a party with so much realistic bass.
 

icerock

Member
Never understood the fascination with comparing console GPU to their PC equivalent, PS4 is basically a cut-down HD7870 and it can run most modern games at a significantly better quality/res/frame-rate compared to a stock HD7870 (which btw is more powerful than PS4 GPU). Just take a look at what games like RDR2, COD:MW etc. look like on HD7870 and then cast a look at base PS4 gameplay, the difference is night and day.

Consoles will always outperform their PC equivalent GPUs because of lower-level APIs and a particular game code being optimised far better for the GPU. So, it was bizarre when Dictator was championing RTX 2060 being the equivalent of PS5 GPU, and his own tests show it to be 20% less performant on AC:Valhalla. I'm sure, over time, these consoles will occasionally out-perform the current equivalent which he is championing (RTX 2080, RTX 3060Ti). Simply, due to nature of game development. You can always squeeze more performance out of a weaker GPU, if you have more time and budget to cater to it.

Those ultra-realistic demos which you see on PC from time to time is a good example, once a dev can work purely concentrate on a single spec. The results are truly ground-breaking. Rebirth demo which runs at 60fps on a single 1080Ti is mind-boggling but also gives you an idea about complex nature of game-development.
 

M1chl

Currently Gif and Meme Champion
Going by the recent Dirt5 analysis. We need to compare more than the FPS at the time , need to check other differences on screen like average-resolution geometry asset-quality shadows anisotropic-filtering ambient-occlusion and much more etc. So it's not just apples to apples comparisons anymore for the same game on both consoles
Yeah VRS makes this more challenging to do, these comparisons I mean.
 

Neo Blaster

Member
Is there a garantee that a more powerful CPU won't affect the results by a adding a couple of FPS even if the scene is GPU bound? Why not use a PS5/XSX equivalent CPU then?
Problem here is that next gen consoles have built-in hardware to perform tasks that CPU would do otherwise, even more on PS5. How can one measure how many core/threads would be necessary to replace that hardware without giving them advantage?
 
Last edited:

Bo_Hazem

Banned
Never understood the fascination with comparing console GPU to their PC equivalent, PS4 is basically a cut-down HD7870 and it can run most modern games at a significantly better quality/res/frame-rate compared to a stock HD7870 (which btw is more powerful than PS4 GPU). Just take a look at what games like RDR2, COD:MW etc. look like on HD7870 and then cast a look at base PS4 gameplay, the difference is night and day.

Consoles will always outperform their PC equivalent GPUs because of lower-level APIs and a particular game code being optimised far better for the GPU. So, it was bizarre when Dictator was championing RTX 2060 being the equivalent of PS5 GPU, and his own tests show it to be 20% less performant on AC:Valhalla. I'm sure, over time, these consoles will occasionally out-perform the current equivalent which he is championing (RTX 2080, RTX 3060Ti). Simply, due to nature of game development. You can always squeeze more performance out of a weaker GPU, if you have more time and budget to cater to it.

Those ultra-realistic demos which you see on PC from time to time is a good example, once a dev can work purely concentrate on a single spec. The results are truly ground-breaking. Rebirth demo which runs at 60fps on a single 1080Ti is mind-boggling but also gives you an idea about complex nature of game-development.

1080p@60fps though. But you don't need too much power with UE5 Nanite and Sony Atom View polygon streaming per frame budget. Death Stranding already looks photorealistic on PS4 Pro, so expect a major jump from there with raytracing as well used smartly.
 

Thirty7ven

Banned
Funniest bit about all the butt-hurt reactions to peformance H2H, is that we are at the least interesting stage of the generation. Once we start seeing true next gen iterations of these engines, and they finally start stressing the CPU and I/O, we will be seeing what the machine was truly built for.
 

Lysandros

Member
Problem here is that next gen consoles have built-in hardware to perform tasks that CPU would do otherwise, even more on PS5. How can one measure how many core/threads would be necessary to replace that hardware without giving them advantage?
I get it, good point. But i doubt that the game's audio and I/O requirements on this specific cut scene justifies the use of that much powerful CPU. A Zen 2 with similar thread count running at 10% or 20% higher frequency would be a fairer comparison in my opinion.
 

Bo_Hazem

Banned
Ok finished WD:L, and I give it 9/10, or even 10/10 for nailing the variety of NPC's/playable characters with impressive sense of humor and character differentiation from each.

I mostly played as an African spy sounds like the presenter here, makes everything funny with his specific sense of humor that sounds serious sometimes. :lollipop_tears_of_joy:





Another more elegantly spoken British spy sounds more "classy", even his sarcasm is classy. But man, can't get enough of that African spy, sounds and acts like a University professor.:lollipop_tears_of_joy:

Sounds are so fucking wonderful in the game, music is just crazy good and the locality is wonderful of 3D audio. Bass is so satisfying with Pulse 3D, and when entering a party you really feel like real sense of party even though overall I think the game doesn't use enough sources to make full 3D audio space, but still so satisfying. Stormzy makes it hard to get out of the car.:lollipop_raising_hand:

The story is brilliant, not dramatic, but more political and convincing.

This concept of NPC's being playable makes me so extremely cautious to not hit any of them by driving fast, something I've never ever experienced before. You might kill a friend of someone even in your clan, a wife, husband, relative, etc.

I got the game for around $41 but bought the season pass after playing it for 3-4 days. Finished most of the game now and I'll be downloading Ass Creed while finishing some very few side missions on WD:L.

See you later!
 

geordiemp

Member
Ok finished WD:L, and I give it 9/10, or even 10/10 for nailing the variety of NPC's/playable characters with impressive sense of humor and character differentiation from each.

I mostly played as an African spy sounds like the presenter here, makes everything funny with his specific sense of humor that sounds serious sometimes. :lollipop_tears_of_joy:





Another more elegantly spoken British spy sounds more "classy", even his sarcasm is classy. But man, can't get enough of that African spy, sounds and acts like a University professor.:lollipop_tears_of_joy:

Sounds are so fucking wonderful in the game, music is just crazy good and the locality is wonderful of 3D audio. Bass is so satisfying with Pulse 3D, and when entering a party you really feel like real sense of party even though overall I think the game doesn't use enough sources to make full 3D audio space, but still so satisfying. Stormzy makes it hard to get out of the car.:lollipop_raising_hand:

The story is brilliant, not dramatic, but more political and convincing.

This concept of NPC's being playable makes me so extremely cautious to not hit any of them by driving fast, something I've never ever experienced before. You might kill a friend of someone even in your clan, a wife, husband, relative, etc.

I got the game for around $41 but bought the season pass after playing it for 3-4 days. Finished most of the game now and I'll be downloading Ass Creed while finishing some very few side missions on WD:L.

See you later!


WTF is that Bo ? I am going to penetrate you witha cucumber !

I watched a few minutes, I will never get them back.
 
Last edited:

FrankWza

Member
That fact that you are even asking this question, means you probably didn't pay attention to past console releases.

I think you missed the point of what I was asking. But PS3 was $200 more than 360 and x1 was $100 more than PS4. And nobody was paying 3 times the price on eBay for any of those systems.
But what we know today is, the PS5 is the better designed console that runs 3rd party multiplat better and has an insane exclusive library and possibly has custom rdna features as well as comparing favorably to a pretty expensive PC. I was asking what it could have sold for at a “fair” price now that we have all this information in front of us.
 

jroc74

Phone reception is more important to me than human rights

Consoles will always outperform their PC equivalent GPUs because of lower-level APIs and a particular game code being optimised far better for the GPU. So, it was bizarre when Dictator was championing RTX 2060 being the equivalent of PS5 GPU, and his own tests show it to be 20% less performant on AC:Valhalla. I'm sure, over time, these consoles will occasionally out-perform the current equivalent which he is championing (RTX 2080, RTX 3060Ti). Simply, due to nature of game development. You can always squeeze more performance out of a weaker GPU, if you have more time and budget to cater to it.
The 2nd part of that quote is interesting....

I always thought it was odd that when RDNA 2 was announced, some ppl kept banging on about the PS5 being a 5700, 5700XT equivalent.

That gpu maxes out at 1440p. I dont even think it can be overclocked as high as a PS5. I agree with those saying the 5700 to gauge PS5 performance was a flawed test when it was done. RDNA 1, RDNA 2 came so fast.....I remember many of us being blindsided when RDNA 1 was announced in a AMD stream thread.

I even asked a few months ago, now that we know RDNA 2 is a thing....why not shelf the RDNA 1 gpu comparisons and wait for an RDNA 2 gpu? Of course some didnt due to a certain narrative, but I thought those that do tech related articles would.
 

sircaw

Banned
Please no.

Side note how I spent my day.....



gQmMGQw.jpg

Why do you torment me so, i hope you fish up one of those old world war 2 mines with the spikes on them.

Or a blue whale who lands on your lap.

and why so many rods ffs, you should be like the old man and the sea throwing out tackle with you bare hands and pulling it in with rags wrapped around your fingers, What kind of sports man are you?

also i bet your one of the those druggie, white trash caravan meth makes and that ocean is actually the ozarks and your real name is Marty. I watched the series!

AND I AM NOT FUCKING ANGRY OR ENRAGED AT ALL.

Looks like you had a nice time outt there. :messenger_heart:
 
Last edited:

Neo Blaster

Member
It does make me wonder if Microsoft was better off raising the Xbox Series X price a little more.

Because PS5 and Series X have price parity, it makes it more justified for people to make direct comparisons. It just seems like it was not an accident that Xbox original and xboxOne were launched at a higher price than the equivalent PS. And the only reason 360 was cheaper than PS3 was because of The Cell.

As it stands, with history as evidence, it does look like that Sony makes more powerful machines for less money than Microsoft.

The suspicion that the heart of Series X was meant to work in servers, also seem to reflect the mistake Sony already made for the PS3. The Cell was meant to be used for everything, and ended up not being that good for running games.
Good remind, Sony indeed wanted to spread Cell everywhere, I remember a tv model was going to use it, just not sure if it came to happen.
 
Status
Not open for further replies.
Top Bottom