Support NeoGAF

[FrankWza]

Especially when the best performing version is Series S, looks like The Coalition did a great job.

oh. But….

Nope. The demo is so variable and jumpy that saying anything about performance is almost useless.

 

[onQ123]

It's funny that he used Outrun in this video because when I think about the new consoles one of the things I always think about is devs using old tricks like they used to with games like Outrun but with the new tech.

 

[AnotherOne]

For old times sake

 

[MasterCornholio]

Its not a broken tools argument.
Its the fact that EPIC have had more time with the PS5 hardware.

Maybe it is higher clocks maybe it isn't.. a pointless discussion, the only info we have on why UE5 is running better on PS5 is the fact epic have had more time with it, not saying thats the reason, but thats the info we have.

Your saying it isn't the reason but you focus heavily on it. Are you blaming it on poor optimization on Epics part?

Anyways I don't see how the PS5s strengths are pointless.

 

[adamsapple]

oh. But….

You're quoting two different people .. what are we supposed to see here exactly ?

 

[Dream-Knife]

So why is PS5 outperforming series x here? Both are on the same architecture, with the series X having a more powerful version. Is it due to some way the series x runs games with that windows subsystem?

 

[FrankWza]

You're quoting two different people .. what are we supposed to see here exactly ?

Don’t THUMB your nose at my post.

https://www.neogaf.com/threads/ignxgamer-matrix-awakens-hellblade-and-the-power-of-unreal-engine-5-performance-preview.1626833/post-265294420

 

[adamsapple]

Don’t THUMB your nose at my post.

https://www.neogaf.com/threads/ignxgamer-matrix-awakens-hellblade-and-the-power-of-unreal-engine-5-performance-preview.1626833/post-265294420

Some of y'all need a break

 

[MasterCornholio]

So why is PS5 outperforming series x here? Both are on the same architecture, with the series X having a more powerful version. Is it due to some way the series x runs games with that windows subsystem?

More powerful in some ways. Which means in certain situations the PS5 will pull ahead depending on what the engine needs.

 

[MonarchJT]

suprising ps5 perf advantage and defenitly bigger than 1fps suggested by df

hahaahaah

 

[Sosokrates]

Your saying it isn't the reason but you focus heavily on it. Are you blaming it on poor optimization on Epics part?

Anyways I don't see how the PS5s strengths are pointless.

The only thing im focusing on is that we dont know why the ps5 is performing better, but the only info we know is that epic got ps5 dev kits first and have had more experience putting UE5 on the ps5.

This is just the reality.

 

[MasterCornholio]

The only thing im focusing on is that we dont know why the ps5 is performing better, but the only info we know is that epic got ps5 dev kits first and have had more experience putting UE5 on the ps5.

This is just the reality.

And that each system has their own strengths. So it can be down to a multitude of reasons.

Edit: To be clear I don't believe there's a massive difference between these systems so I'm not really surprised at these results. The two are extremely close and I would consider it a tie. The only argument that I'm not buying is the broken tools one especially a year after these systems are out.

 

[Sosokrates]

I

And that each system has their own strengths. So it can be down to a multitude of reasons.

Ive already said this.

But we have no info leading to any reason at the moment.

I mean people can say its lack of optimization or higher ps5 clock speed until the sun goes down, but its just speculation with nothing pointing towards it.

 

[onQ123]

Maybe it's the higher clocks then if it isn't the SSD.

I don't believe in the broken tools argument because the system as been out a while. Some of the PS5s strengths helping it out in certain situations makes more sense to me.

If the Demo was released a year ago that extra time would have played a big role but now that small extra time with PS5 dev kits is nothing

So why is PS5 outperforming series x here? Both are on the same architecture, with the series X having a more powerful version. Is it due to some way the series x runs games with that windows subsystem?

Different designs games will always hit different bottlenecks on these systems .

 

[Werewolfgrandma]

The only thing im focusing on is that we dont know why the ps5 is performing better, but the only info we know is that epic got ps5 dev kits first and have had more experience putting UE5 on the ps5.

This is just the reality.

My assumption is that this team at epic now has had the better part of 2 years developing for ps5 and 50% or less for Xbox. Seems pretty reasonable for any differences.

 

[onQ123]

My assumption is that this team at epic now has had the better part of 2 years developing for ps5 and 50% or less for Xbox. Seems pretty reasonable for any differences.

So you think Epic didn't touch a Xbox Series Devkit until the console was released?

 

[Sosokrates]

My assumption is that this team at epic now has had the better part of 2 years developing for ps5 and 50% or less for Xbox. Seems pretty reasonable for any differences.

Not according to onQ123

 

[MasterCornholio]

I

Ive already said this.

But we have no info leading to any reason at the moment.

I mean people can say its lack of optimization or higher ps5 clock speed until the sun goes down, but its just speculation with nothing pointing towards it.

I believe O onQ123 has a point about the dev time difference between small between the two. I'm not a developer myself so I'm not sure how much optimizations can be done in such a short period of time. It doesn't appear that one version is horribly optimized compared to the other.

 

[DeepEnigma]

Not according to onQ123

What is a Fortnite? What is an engine MS heavily uses? What is an engine that Epic wants every developer to use thus they are working tirelessly with devkits of all platforms to make sure it scales well to them?

 

[MasterCornholio]

So you think Epic didn't touch a Xbox Series Devkit until the console was released?

Oh hell no. Microsoft would be stupid to not send Epic some dev kits before the system released. Especially since Epic was developing a next gen engine that many of Microsofts 1st party studios use.

 

[DeepEnigma]

Oh hell no. Microsoft would be stupid to not send Epic some dev kits before the system released. Especially since Epic was developing a next gen engine that many of Microsofts 1st party studios use.

 

[Sosokrates]

I believe O onQ123 has a point about the dev time difference between small between the two. I'm not a developer myself so I'm not sure how much optimizations can be done in such a short period of time. It doesn't appear that one version is horribly optimized compared to the other.

Well I dont know why one would think that when EPIC made the valley demo for the PS5 and didn't for the seriesX because they didnt have the dev kits, doesn't seem an insignificant amount of time to me.

 

[onQ123]

Not according to onQ123

50% of 2 years is 1 year so the only way that Epic could have had 50% or less time with Xbox Series devkits is that if they didn't start working with Xbox Devkits until after the console was released.

 

[Loxus]

The only thing im focusing on is that we dont know why the ps5 is performing better, but the only info we know is that epic got ps5 dev kits first and have had more experience putting UE5 on the ps5.

This is just the reality.

That mainly only applies to launch day releases.

Series console been out for over a year. Epic had enough time to get used to the hardware and even have Nanite using Mesh Shader.

 

[Sosokrates]

50% of 2 years is 1 year so the only way that Epic could have had 50% or less time with Xbox Series devkits is that if they didn't start working with Xbox Devkits until after the console was released.

Dude... We dont even know when epic got dev kits. But they said they didn't do the valley demo on xsx because they didnt get dev kits, so it must of been quite significant.

 

[DeepEnigma]

 

[onQ123]

Dude... We dont even know when epic got dev kits. But they said they didn't do the valley demo on xsx because they didnt get dev kits, so it must of been quite significant.

You clearly over looked the point of my post . He said they had 50% of the time with the Xbox Series devkit & that they had 2 years with PS5 devkit. that would mean that Epic didn't get a Series devkit until after the console was released.

 

[MasterCornholio]

Dude... We dont even know when epic got dev kits. But they said they didn't do the valley demo on xsx because they didnt get dev kits, so it must of been quite significant.

Your just assuming that Epic got the dev kits a lot later. I mean why would they have gotten the dev kits 6 months later?

 

[adamsapple]

How does a video from NXGamer about UE5 as a piece of tech turn into a debate about which console runs a technical demo, not indicative of any retail game, better..

The fact that they had to work with Coalition to not only get the demo up for the Series consoles but also to take their help in improving the engine as a whole shows the engine is still very much a work-in-progress on consoles.

STALKER 2 is gonna be a good first taste of what UE5 on consoles can do. Gonna be interesting to see if it uses the recently shown UE5 bells and whistles.

 

[Sosokrates]

You clearly over looked the point of my post . He said they had 50% of the time with the Xbox Series devkit & that they had 2 years with PS5 devkit. that would mean that Epic didn't get a Series devkit until after the console was released.

No, they could of got PS5 devkits jan 2019 and xsx devkits jan 2020.

 

[adamsapple]

Your just assuming that Epic got the dev kits a lot later. I mean why would they have gotten the dev kits 6 months later?

While we don't know how late, we know for a fact that MS sent out Series console dev kits later than Sony did for PS5.

So, naturally, more developers had a head start on PS5 development. Again, we can't say exactly how long it is.

But there is definitely a noteworthy gap between these two dev kits being sent out.

Xbox Series X/S Performance Issues Due In Part To Dev Kits Coming Later Than PS5 Kits, Claims New Report

gamingbolt.com

 

[onQ123]

No, they could of got PS5 devkits jan 2019 and xsx devkits jan 2020.

That would be closer to 70% of the time

 

[sircaw]

How does a video from NXGamer about UE5 as a piece of tech turn into a debate about which console runs a technical demo, not indicative of any retail game, better..

The fact that they had to work with Coalition to not only get the demo up for the Series consoles but also to take their help in improving the engine as a whole shows the engine is still very much a work-in-progress on consoles.

STALKER 2 is gonna be a good first taste of what UE5 on consoles can do. Gonna be interesting to see if it uses the recently shown UE5 bells and whistles.

i really hope its 60 fps, not too keen on shooters being 30.( talking about Stalker ofc)

 

[Physiognomonics]

So why is PS5 outperforming series x here? Both are on the same architecture, with the series X having a more powerful version. Is it due to some way the series x runs games with that windows subsystem?

Your error is to believe that narrative setup mainly by DF before those machines were released. You'll notice they changed their tune and usually don't talk anymore about 12TF being better than 10TF.

 

[kingfey]

I will wait for stalker 2.

That game will test, whether the new consoles are ready for true pc next gen game, which utilizes UE5.

 

[SlimySnake]

I will wait for stalker 2.

That game will test, whether the new consoles are ready for true pc next gen game, which utilizes UE5.

Pretty sure Stalker 2 isnt using lumens or Nanite.

 

[adamsapple]

i really hope its 60 fps, not too keen on shooters being 30.( talking about Stalker ofc)

We know they're supporting 4K, 60 FPS and RT.

Whether all 3 are the same mode like Metro or separate is not yet known.

 

[Negotiator101]

50% of 2 years is 1 year so the only way that Epic could have had 50% or less time with Xbox Series devkits is that if they didn't start working with Xbox Devkits until after the console was rereleased.

Is this really a discussion worth having?

 

[MasterCornholio]

While we don't know how late, we know for a fact that MS sent out Series console dev kits later than Sony did for PS5.

So, naturally, more developers had a head start on PS5 development. Again, we can't say exactly how long it is.

But there is definitely a noteworthy gap between these two dev kits being sent out.

Xbox Series X/S Performance Issues Due In Part To Dev Kits Coming Later Than PS5 Kits, Claims New Report

gamingbolt.com

I think it's safe to say that Epic had both def kits for a while. Pretty sure they spent a lot of time with each one for this demo.

 

[VFXVeteran]

I come from the 3D animation and VFX world, use tons of different softwares today, and even more thoughout the years.

Naming conventions change from software to software, one calls it potato other calls it onion, they both do the same thing.

Calling it a PROXY is not incorrect, because it's indeed a PROXY, even tho inside UE5 it means another thing or has another context to it.

I stand by @NXGamer on ths one. Let's not make a big deal out of it.

Yes, I agree with you on this but giblet did indeed describe the actual underpinnings of what's going on at the data structure level. I wouldn't have come down hard claiming each doesn't have any knowledge on the subject though. That's a little trollish imo. I get NXGamer's context and I also get the point about being more detail-oriented on the actual implementation details of what Nanite is (which, to be fair, is probably overkill for most people anyway).

I liken Nanite to MIP-mapping in a way. While the data for all the levels are in memory at once, you could say each level is a "proxy" of the lowest level mip.

 

[John Wick]

We dont know why.
Epic have confirmed they have had UE5 on PS5 for longer though, or UE5 could be more suited to PS5's hardware/dev environment.
But we have no information pointing to any of these reasons, so pushing one or more of these reasons shows a bias or intention to push a narritive.

It's pointless too. The code isn't optimised anyway. It's all work in progress.

 

[mortal]

Gamers today are fucking spoiled lol

 

[adamsapple]

I think it's safe to say that Epic had both def kits for a while. Pretty sure they spent a lot of time with each one for this demo.

Unless most of the Series version development was off-loaded to The Coalition. We can never say anything about these kind of things for sure, and neither Epic nor Coalition are ever gonna fully talk about it.

 

[DeepEnigma]

Is this really a discussion worth having?

He didn't bring it up.

 

[Werewolfgrandma]

So you think Epic didn't touch a Xbox Series Devkit until the console was released?

Why go straight to crazy? This team specifically made the first ue5 ps5 demo...I think. Giving this team the team that made the demo much more experience on ps5.

 

[SlimySnake]

So why is PS5 outperforming series x here? Both are on the same architecture, with the series X having a more powerful version. Is it due to some way the series x runs games with that windows subsystem?

Same RDNA2.0 architecture but not same overall architecture. The PS5 I/O could be helping the PS5 GPU overperform its tflops here.

Also, the RDNA 2.0 architecture relies on really high clocks up to 2.7 GHz to hit its performance targets. The PS5 is at 2.23 Ghz while the XSX is even lower at 1.8 Ghz. It's possible that the lower clocks are holding back the 52 CU XSX GPU.

Lastly, the XSX uses a RDNA 2.0 chip that adds more CUs to a 2 Shader Array system which is probably causing some kind of bottleneck where the CUs arent being effectively utilized. The 13 tflops AMD 6700xt does not use 52 CUs. It tops out at 40 CUs and pushes the clocks up to 2.4 Ghz to hit its tflops target. It seems even AMD knew that the best way to get performance out of that particular CU configuration. This is also something Cerny hinted at in his Road to PS5 conference. Something we initially dismissed as damage control.

Another potential difference between the XSX and RDNA 2.0 PC cards is that it lacks the infinity cache thats part of the GPU die. The 6700xt is a 337 mm2 GPU compared to the entire XSX APU which includes the CPU and IO and still comes in around 360 mm2. So how can a 40 CU GPU be almost as big as the entire 52 CU XSX APU? The infinity cache must be taking up a lot of that space, and thus must be the reason why the XSX might not be performing up to its tflops potential.

 

[adamsapple]

Why go straight to crazy? This team specifically made the first ue5 ps5 demo...I think. Giving this team the team that made the demo much more experience on ps5.

There's a lot of room between 0 and 100.

O onQ123 we don't have to jump to either extremes lol

 

[isoRhythm]

Alex going out of his way to get into a slow ass SUV on the xsx was suspect from the beginning. Maybe heisenberg was right after all.

I'm not big on conspiracies or seeing DF as purposely doing something like that. But the differences found by Kingthrash (Yes ik he's a warrior, doesn't negate his findings) and now by NX as well point to DF either outright avoiding the differences or being incompetent in finding them. Not a good look.

 

[Cherrypepsi]

 

[Negotiator101]

Same RDNA2.0 architecture but not same overall architecture. The PS5 I/O could be helping the PS5 GPU overperform its tflops here.

Also, the RDNA 2.0 architecture relies on really high clocks up to 2.7 GHz to hit its performance targets. The PS5 is at 2.23 Ghz while the XSX is even lower at 1.8 Ghz. It's possible that the lower clocks are holding back the 52 CU XSX GPU.

Lastly, the XSX uses a RDNA 2.0 chip that adds more CUs to a 2 Shader Array system which is probably causing some kind of bottleneck where the CUs arent being effectively utilized. The 13 tflops AMD 6700xt does not use 52 CUs. It tops out at 40 CUs and pushes the clocks up to 2.4 Ghz to hit its tflops target. It seems even AMD knew that the best way to get performance out of that particular CU configuration. This is also something Cerny hinted at in his Road to PS5 conference. Something we initially dismissed as damage control.

Another potential difference between the XSX and RDNA 2.0 PC cards is that it lacks the infinity cache thats part of the GPU die. The 6700xt is a 337 mm2 GPU compared to the entire XSX APU which includes the CPU and IO and still comes in around 360 mm2. So how can a 40 CU GPU be almost as big as the entire 52 CU XSX APU? The infinity cache must be taking up a lot of that space, and thus must be the reason why the XSX might not be performing up to its tflops popotential.

You had to ask