Current Generation Hardware Speculation with a Technical Spin [post GDC 2020] [XBSX, PS5]

[Tkumpathenurple]

My guess would be that right now the PS5 is saving more bandwidth on stuff like asset loading (meshes, textures, maybe even sounds, anything) than the Xbox can make up for in its faster 10GB of RAM. That is probably slightly more likely than that the remaining 3.5GB is slower and causing issues there what I thought initially, but you never know for sure in these early days. It could even be both, or neither, but the explosions do suggest bandwidth? So would the absense of the smoke from the gun on Xbox if that was deliberate.

My guess is that it's a combination. Remember the slide from a PS4 presentation, stating that CPU access effected the GPU bandwidth disproportionately? I reckon that something similar is afoot with the slower 3.5GB's, and Microsoft's dev tools aren't yet mature enough to mitigate/nullify that.

Is next year's GDC the earliest point at which we stand a chance of seeing some leaked slides from multiplatform developers?

 

[Deleted member 7537]

So, after months of:

• “PS5 is really a 9TF console or even lower because frequencies will drop most of the time”
• “The difference is 20% at best but probably much higher than that”
• “Sony upclocked the PS5 after panicking”
• “Multiplat games will run at 4k60 on XBX and at 1440p30 on PS5”

It turns out that they were actually trolling and hadn’t got a clue.

 

[Strange]

Kind of funny how after months of slamming PS5, in world results are that they're really in the same ballpark- much too close for comfort for Microsoft.
Unlike how PS4 had the edge over XBO hands down across the board, they're really trading blows with PS5 apparently lower specced on paper.

 

[Globalisateur]

It's time people start to acknowledge the advantage of fast and narrow design vs wide and slow (this includes number of CU by SE). It doesn't explain all, but it's not only about tools. This is what Matt from resetera said about it. Matt is a known and trusty multiplat developer:

1) It’s early in the gen on brand new hardware, it will take time for devs to get fully up to speed
2) The PS5 dev tools are simply better and more mature right now
3) It turns out Sony wasn’t lying about “fast and narrow” having some benefits over “wide and slow”

 

[Karamazov]

still too early to draw conclusions.
The first "real" next gen comparison may be Cyberpunk.
Still it's nice to see the PS5 doing well for now. Can't wait to get mine at the end of the week hopefully !

 

[Inuhanyou]

Im glad both machines are in the same ballpark. I expect as apis improve for series x to improve but i hope not by much. Its less fan wars if the two systems can compete on things other than raw number and who has more powerful box

 

[AbsoluteBeginner]

PS5 reminds me of 360, has all the advantage except for theoretical power.

I do suspect XSX will soon have advantage, albeit slight, but anything close will be seen as win for PS5 (and for a good reason, it has advantages in other areas).

Dont forget that Sony delivered their dev kits with actual SOC back in June 2019, long before MS sent theirs.

It's time people start to acknowledge the advantage of fast and narrow design vs wide and slow (this includes number of CU by SE). It doesn't explain all, but it's not only about tools. This is what Matt from resetera said about it. Matt is a known and trusty multiplat developer:

Sonys "fast and narrow" definitely has advantage when fast is that much faster ~400MHz and ROPs and front end are same for both.

 

[Tkumpathenurple]

It's time people start to acknowledge the advantage of fast and narrow design vs wide and slow (this includes number of CU by SE). It doesn't explain all, but it's not only about tools. This is what Matt from resetera said about it. Matt is a known and trusty multiplat developer:

It's already known that there are some advantages to fast & narrow vs slow & wide, but it's a bit premature to properly state what those are, as we've not seen this kind of comparison before.

The PS4 (18CU @ 800MHz) vs the XBoxOne (12CU @ 853MHz) were so similar in terms of clockspeed that the most substantial difference came from CU count and easily usable bandwidth.

The PS4 Pro (36CU @ 911MHz) vs the XBoxOne X (40CU @ 1.172GHz) was a mild difference in terms of CU count and a notable difference in terms of clockspeeds and bandwidth. All in favour of the X1X.

This is the first time that we're seeing consoles of the same architecture battling it out with one being much faster, but the other much larger. Also, while the PS5 has a more straightforward UMA, the XSX has higher bandwidth for 10GB's.

Once tools are matured in a year or two, we'll have a much better idea of which advantage leads to which benefits.

It'll be interesting to see - and results are currently so similar that I hope fanboys shut their sad, little traps - but I don't think we can call much of any advantage yet, beyond well developed tools. That Sony have gone from the travesty of the PS3 to having the best tools (out of the gate, at least) for two generations on the trot now is, I think, testament to Mark Cerny's philosophy.

 

[Tabris]

The RAM inside PS5 D9XKV seems a newer customized version
https://www.4gamer.net/games/990/G999027/20201116053/SS/023.jpg

 

[AlNom]

Looking at the video the only real difference is in the 120fps mode, and it almost always is during heavy explosions with smoke effects where the Xbox drops a little lower although on 120fps 5fps difference isn’t much.

Despite the RAM bandwidth differential, there is still bandwidth compression in the ROPs plus the higher clock rate to contend with.

 

[AbsoluteBeginner]

Could point to a ROP difference that we haven't publicly heard about.

Despite the RAM bandwidth differential, there is still bandwidth compression in the ROPs plus the higher clock rate to contend with.

What is that we havent publicly heard about? Both have 64ROPs, but PS5 fillrate is ~25% higher.

Aha I see edit, yes.

 

[AlNom]

What is that we havent publicly heard about? Both have 64ROPs, but PS5 fillrate is ~25% higher.

Aha I see edit, yes.

Right, I'm not sure if it was publicly announced what the ROP counts were, but there is certainly the clock differential that could make up for any hiccoughs seen in ROP-heavy scenes in spite of raw bandwidth. Hard to say without knowing more.

 

[rekator]

Despite the RAM bandwidth differential, there is still bandwidth compression in the ROPs plus the higher clock rate to contend with.

So the devil is in the detail…

 

[London Geezer]

So, after months of:

• “PS5 is really a 9TF console or even lower because frequencies will drop most of the time”
• “The difference is 20% at best but probably much higher than that”
• “Sony upclocked the PS5 after panicking”
• “Multiplat games will run at 4k60 on XBX and at 1440p30 on PS5”

It turns out that they were actually trolling and hadn’t got a clue.

Well, whoever said those things was obviously trolling even before we saw these results, as 4k60 v 1440p30 given the difference on paper is fantasy. But yeah, I didn't expect the PS5 to keep up quite as much as this.

 

[iroboto]

Right, I'm not sure if it was publicly announced what the ROP counts were, but there is certainly the clock differential that could make up for any hiccoughs seen in ROP-heavy scenes in spite of raw bandwidth. Hard to say without knowing more.

There’s a lot happening at once for launch, without more data points it going to be hard to figure out which areas could be a bottleneck.

though I suspect this plays a role somewhere in there.

 

[iroboto]

Well, whoever said those things was obviously trolling even before we saw these results, as 4k60 v 1440p30 given the difference on paper is fantasy. But yeah, I didn't expect the PS5 to keep up quite as much as this.

Sounds like a statement either taken out of context or mixed contexts between two separate posts.

I’d challenge him to find that statement where someone would suggest a 4x pixel difference for all multiplats. I’ve never seen it.

 

[Deleted member 7537]

Sounds like a statement either taken out of context or mixed contexts between two separate posts.

I’d challenge him to find that statement where someone would suggest a 4x pixel difference for all multiplats. I’ve never seen it.

Lol, if I remember correctly you were the one saying PS5 would render mp games at 1440p while XBSX would render them at 4K. 120% difference in performance. "I've never seen it".

 

[iroboto]

Lol, if I remember correctly you were the one saying PS5 would render mp games at 1440p while XBSX would render them at 4K. 120% difference in performance. "I've never seen it".

Find it. You’ve taken my words out of context. Pretty sure I was speaking about anomalous situations for frame rate cutoffs which has nothing to do with hardware performance. In the same post I spoke that there will be situations that PS5 will outperform XSX and on the extreme other setup such an event could happen.

There are far too many regulars here that converse with me to have not called me out on that.

I most certainly did not say 1440p30 vs 4K60 in that post either. That’s a 400% increase. And I most certainly did not say all multiplats.

Find the post and read it again and following conversation of that and you’ll see you’re wrong.

 

[Deleted member 7537]

Nah, you go find it. After months of reading all sorts of fud and concern trolling and now that we see the real results, there isn't really much to discuss. PS5 is a damn fine console, the issue is with people that want their console of choice to do better and the other ones to do poorly, even if it's only on their imagination.

 

[AbsoluteBeginner]

Nah, you go find it. After months of reading all sorts of fud and concern trolling and now that we see the real results, there isn't really much to discuss. PS5 is a damn fine console, the issue is with people that want their console of choice to do better and the other ones to do poorly, even if it's only on their imagination.

How about you find it since you are the one that brought this claim to the conversation? I can guarantee no body sane here expected 2TF difference to result in 2.2x more pixels pushed to the screen, especially considering BW/ROPs of both consoles. If this was said by twitter weirdos then take it there, dont derail the thread here.