• 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.

Platform Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block

Feb 5, 2019
283
465
255
And this is just ( one of ) the reason why i had right to doubt about Discord. Even members of ordinary Xbox Discord are doing that. Continuously spreading FUD about PS5 is astonishing
Again that post AND THIS thread is about XVA, SFS etc...in the XSX. Not about PS5 directly. Only PS5 mentions are for legitimate comparisons.
 
Last edited:

DarthFeanor

Member
May 7, 2020
100
104
185
I recommend people take a look at Digital Foundry discussion about Senua Saga trailer back in December.
They were impressed by the level of environmental details and if that was possible in real time game play.
It seems it is possible with Nanites or something like that, even though Senua Saga is probably using Unreal 4.
 

THEAP99

Member
Mar 16, 2020
774
1,285
370
here was this whole drama in Meta GAF, where the THEAP99 THEAP99 THEAP99 THEAP99 and his fans tried to get me banned from GAF first with out of context screenshot from my discord then with a photoshop of me being admin in XboxEra.
I didn't try getting anyone banned from anything. i only did what i was told to do so that I would be cleared from whatever drama was going to take place
 

Redlight

Member
Feb 5, 2014
215
507
465
lol no it can run on XSX, PC and laptops too (even better in some cases) .... your buying into the sony marketing crap.....
Just for the sake of clarity. The partial quote you're responding to was part of a longer argument pointing out that Epic have never made the claim that the Unreal 5 demo couldn't be done on the Series X. :)
 

X-Fighter

Member
Jan 10, 2020
1,823
2,543
355
Interesting. I hadn't seen it yet. When he says:

If the speed of texture loading, after decompression, is around 3.8 GiB/s, that would allow loading 64 MiB per frame at 60fps. This is fast enough to load 1000 tiles each frame, about the equivalent of a single 4096x4096 texture.
How many tiles are their normally in a frame?
 
  • Like
Reactions: FALCON_KICK

Bernkastel

Formerly 'Lady Bernkastel'
Mar 8, 2018
3,025
5,396
840
Ljubomir Peklar, designer at Ebb Software
Series X is a very balanced system. The SSD solution(referring to Xbox Velocity Architecture) is an incredible improvement when it comes to loading assets, and as it stands now it seems that there will be complete parity between PC and Series X versions of the game.
 
Last edited:

MrMiyagi

Member
Mar 2, 2020
85
56
160
I dunno but so far we've seen State of Decay in 10 seconds on Series X and Spider man loading instantly on ps5.
 

Bernkastel

Formerly 'Lady Bernkastel'
Mar 8, 2018
3,025
5,396
840
I dunno but so far we've seen State of Decay in 10 seconds on Series X and Spider man loading instantly on ps5.
You realize that Spiderman was optimized for the PS5? None of the games shown for Xbox was optimized. It was just regular quality of life functions just by being played on the machine. Only way to judge is to see how long it takes to go from XSX games.
Not to mention State of Decay 2 has load times from 40 seconds to around a minute, way more than Spiderman.
Why didnt Sony used Days Gone instead ? Xbox used the Xbox One X version of State of Decay 2, while Sony optimized Spiderman for PS5. Ultimately, Xbox took the worst offender of their platform and ran it unoptimized, Sony took the best thing and still optimized it for PS5.
 

sinnergy

Member
Jun 16, 2007
3,381
1,344
1,285
Why didnt Sony used Days Gone instead ? Xbox used the Xbox One X version of State of Decay 2, while Sony optimized Spiderman for PS5. Ultimately, Xbox took the worst offender of their platform and ran it unoptimized, Sony took the best thing and still optimized it for PS5.
I agree, but you could say it’s bad practice for marketing ;) as people like to be wowed. And like the best, So Sony did better in that regard. Now if that reflects real word performance is a whole different matter.
 
Last edited:
Aug 28, 2019
2,539
4,785
540
www.instagram.com
I dunno but so far we've seen State of Decay in 10 seconds on Series X and Spider man loading instantly on ps5.
6.5 seconds, and IIRC they saved the game and then did a load state. Either way, 6.5 seconds falls more or less exactly in line with the 2.4 GB/s raw sequential read speed MS's advertised.

The SpiderMan load was 0.8 seconds. Not instant, though very fast. But it was seemingly for a slice of the game.
 
Last edited:
Nov 5, 2016
23,807
45,534
1,135
One Big Room, Full Of Bad Bitches
I dunno but so far we've seen State of Decay in 10 seconds on Series X and Spider man loading instantly on ps5.
It wasn’t 10 seconds, and it wasn’t even optimized yet. We don’t even know what that Spidey demo fully entailed. Was it the full game?

State of Decay 2 goes from loading in about 45 seconds on Xbox One X to just seven seconds on Xbox Series X. Even more impressively, this is with no extra work or optimization done, that's just how much State of Decay 2 benefits when installed on an Xbox Series X.
 
Mar 27, 2020
3,086
8,299
555
THEAP99 THEAP99


I think you two need to take that stuff here and I'll join you if I think it's worth it.
 
Aug 28, 2019
2,539
4,785
540
www.instagram.com
And they could have done better. Obviously they made a really bad choice with that demo no idea how they came to that decision.

It's up to them to prove that they can do better.
How was it a bad demo? It loaded in 6.5 seconds, right in line with the drive's stated sustained raw sequential read speed. And it still wasn't optimized, at that.

People are blowing the SOD2 thing completely out of proportion and seemingly forgot how to count and what the other numbers already provided were when trying to count the time. Could they have done another game and not do a save > load state? Yes. But it's not like the demo was a disaster, either.
 

SirTerry-T

Member
Dec 4, 2018
385
589
350
U.K
I dunno but so far we've seen State of Decay in 10 seconds on Series X and Spider man loading instantly on ps5.
Ten whole seconds eh? That's practically a lifetime. How will gamers cope?

Hands up who remembers trying...and sometimes failing to load games from a cassette tape?

Hmmm, not many hands going up, thought as much.
 
Mar 27, 2020
3,086
8,299
555
How was it a bad demo? It loaded in 6.5 seconds, right in line with the drive's stated sustained raw sequential read speed. And it still wasn't optimized, at that.

People are blowing the SOD2 thing completely out of proportion and seemingly forgot how to count and what the other numbers already provided were when trying to count the time. Could they have done another game and not do a save > load state? Yes. But it's not like the demo was a disaster, either.
Well Microsoft should have shown a more optimized demo to give people an idea of what the I/Os capabilities are.

That's why called it a bad demo because they could have done a much better job demonstrating the I/O.
 

Trueblakjedi

Banned
Mar 30, 2020
342
715
375
The ps4 Spiderman demo load was 8.10 seconds on a PS4 (im sure that that was NOT the whole game)

The optimized demo loaded in .8 seconds.

The State of Decay demo was a 45 second full game load.

That same game with no optimization loaded in 6.5 seconds. Sure an optimized load would be less... if an optimized load of the same game approached 4.5 seconds (optimization reducing load by 2 seconds) then the ratio is virtually the same.
 
x1

Bernkastel

Formerly 'Lady Bernkastel'
Mar 8, 2018
3,025
5,396
840
And they could have done better. Obviously they made a really bad choice with that demo no idea how they came to that decision.

It's up to them to prove that they can do better.
I guess going with the game that has the biggest loading screens to showcase faster loading times is a very logical choice and not make everything look like a PR.
 
Last edited:
  • Like
Reactions: Trueblakjedi
Mar 27, 2020
3,086
8,299
555
I guess going with the game that has the biggest loading screens to showcase faster loading times is a very logical choice and not make everything look like a PR.
Well I keep hearing that it wasn't optimized. And if it was then we would have gotten much better results.

Are you suggesting that it was optimized?
 

Bernkastel

Formerly 'Lady Bernkastel'
Mar 8, 2018
3,025
5,396
840
Well I keep hearing that it wasn't optimized. And if it was then we would have gotten much better results.

Are you suggesting that it was optimized?
It was not, but would they actually optimize all the games from 3 generations of Xbox? Its better to show what the tools can do
 

Trueblakjedi

Banned
Mar 30, 2020
342
715
375
Can't be that difficult to put together an optimized demo to show off smart delivery and lightning fast load times.

It would have been cool if they did a side by side comparison of the upgraded version of the game versus the vanilla version of it.
Agreed. I think they were trying to show off the raw improvement which a developer would receive before doing anything.

I think that is also a very sound strategy... then everything became a benchmark against the competitor. Unfortunate that thats where we are.

Can't win either way really.
 
  • Like
Reactions: darrenskywalker

Jon Neu

..he neus nothing
Jan 21, 2018
2,710
4,767
675
Republic of Catalonia
And they could have done better. Obviously they made a really bad choice with that demo no idea how they came to that decision.

It's up to them to prove that they can do better.
MS probably didn’t anticipate that them loading Xbox One games in 8 seconds without any optimization would be used by some to attack the console.
 
Mar 27, 2020
3,086
8,299
555
MS probably didn’t anticipate that them loading Xbox One games in 8 seconds without any optimization would be used by some to attack the console.
It was the 1st showing of the load times and they should have optimized the demo better. It's just weird that they didn't optimize it in the first place.
 
Mar 27, 2020
3,086
8,299
555
Its not any Xbox One game though. Its the worst first party offender.
They could have used Gears 5.

That would have been a more impressive demo of the I/O in my opinion.

But instead they chose an unoptimized 1st party title that's kind of a mess. It's just a weird choice in my opinion.
 

Dee_Dee

Member
Apr 22, 2020
250
620
265
Lol they show a game loading in 8 seconds on the SeX vs 40 seconds on the One X WITHOUT BEING OPTIMIZED and yet people still find something to complain about. Y’all are something else.

Luckily they have their hardware talk in August and will hopefully demonstrate what it will be like for games that are actually built to take advantage of XVA.
 

Jon Neu

..he neus nothing
Jan 21, 2018
2,710
4,767
675
Republic of Catalonia
It was the 1st showing of the load times and they should have optimized the demo better. It's just weird that they didn't optimize it in the first place.
They wanted to show how by just playing the old games in the new hardware, you have already a huge advantadge in loading times without any optimization.

I don’t see how that is weird.
 

Ascend

Member
Jul 23, 2018
1,842
2,613
485
It was the 1st showing of the load times and they should have optimized the demo better. It's just weird that they didn't optimize it in the first place.
It was used to show backwards compatibility. And for some reason, no one ever reads the description, which says clear as day that it does not represent a game optimized for the XSX.

Man how the quality of this thread has degraded.
 
Last edited:

geordiemp

Member
Sep 5, 2013
8,615
5,568
790
UK
The ps4 Spiderman demo load was 8.10 seconds on a PS4 (im sure that that was NOT the whole game)

The optimized demo loaded in .8 seconds.

The State of Decay demo was a 45 second full game load.

That same game with no optimization loaded in 6.5 seconds. Sure an optimized load would be less... if an optimized load of the same game approached 4.5 seconds (optimization reducing load by 2 seconds) then the ratio is virtually the same.

Noticed this thread still cannot count to ten yet.
 

Bernkastel

Formerly 'Lady Bernkastel'
Mar 8, 2018
3,025
5,396
840

Noticed this thread still cannot count to ten yet.
A side-by-side load of State of Decay 2 illustrates that on an Xbox One X, it take about 52 seconds to load the open-world game from its main menu. On Xbox Series X, that load time is slashed to about 10 seconds.
Stop wasting our time if you dont have anything meaningful to add.
 
Last edited:
Nov 5, 2016
23,807
45,534
1,135
One Big Room, Full Of Bad Bitches
Why do different articles have different times?

State of Decay 2 goes from loading in about 45 seconds on Xbox One X to just seven seconds on Xbox Series X. Even more impressively, this is with no extra work or optimization done, that's just how much State of Decay 2 benefits when installed on an Xbox Series X.

Spidey demo

This is why, when PlayStation revealed how fast the PlayStation 5 is able to load a level in its game Marvel's Spider-Man - in just one second compared to the PS4 Pro's already-speedy five seconds - players were impressed.

I’ve read an article that said 7 seconds for SoD, I’ve read another that said 11, and everything in between. They also have different times for the One X performance as basis of comparison too
 
Last edited:
  • Like
Reactions: Hendrick's

Bernkastel

Formerly 'Lady Bernkastel'
Mar 8, 2018
3,025
5,396
840
OP updated with who James Stanard, Richard Geldreich and Mike Evans. Link to more Richard Geldreich tweets.
PS : James Stanard is Graphics Optimization R&D and Engine Architect at Microsoft. Richard Geldreich has worked at Space X, Valve and Ensemble. Mike Evans has worked for AMD and NASA.
More and even more from Richard Geldreich on Xbox Series X hardware decompression techniques.
 
Last edited:

Bernkastel

Formerly 'Lady Bernkastel'
Mar 8, 2018
3,025
5,396
840
Man how the quality of this thread has degraded.
I would say it has gotten better. The first 17 pages were an absolute shitshow, with one post(after MiyazakiHatesKojia says devs like PS5 more) where someone reposts a ResetEra post about how many devs have liked PS5 more full of links to out of context tweets and RestEra posts from devs including this from Mike Evans(how are these even relevant to the thread?). After Page 17 people actually started talking about XVA. It definitely got a bit better.
 

Ascend

Member
Jul 23, 2018
1,842
2,613
485
I would say it has gotten better. The first 17 pages were an absolute shitshow, with one post(after MiyazakiHatesKojia says devs like PS5 more) where someone reposts a ResetEra post about how many devs have liked PS5 more full of links to out of context tweets and RestEra posts from devs including this from Mike Evans(how are these even relevant to the thread?). After Page 17 people actually started talking about XVA. It definitely got a bit better.
Fair enough. I was mainly comparing it to the bulk of the discussion that started to happen after page 20 or so. There were some good discussions there. After page 45 it all turned to the same discussions in the PS5/(not)XSX speculation thread.

In any case...

How many tiles are their normally in a frame?
That depends on a LOT of things. It's even hard to give a ball park figure.

If you have a wall, and you're directly in front of it, you might have a bunch of repeating mips, meaning you don't have that many unique tiles. Anything that is repeated doesn't require multiple loadings.
On the other hand, if the wall is extremely large and you're looking at it at an angle, you'd have to load multiple mip levels with different details in order to not overtax the GPU with unnecessary calculations for the mips that are located far away from the camera.

And you're gonna have draw distance be an influence on this, since generally the more draw distance the higher the chance unique assets need to be loaded.
If you have lots of different foliage it's gonna become more taxing... Too many variables.

I guess the absolutely worst case scenario that you will never reach is if each pixel requires a different tile. For practical purposes, all we can really say is that it's way way WAY below that.
I have no idea how many tiles you would need to load on average. But considering Microsoft measured the amount of textures that are actually used compared to the ones that are loaded, they most likely have a very good idea how much tiles require to be loaded each frame, and I don't think they'd deliberately under-design the console.
 
Aug 28, 2019
2,539
4,785
540
www.instagram.com
Why do different articles have different times?




Spidey demo




I’ve read an article that said 7 seconds for SoD, I’ve read another that said 11, and everything in between. They also have different times for the One X performance as basis of comparison too
Lack of consistency and poor counting, that's all. I started getting into keeping track of that stuff myself after the Rose Tico "controversy" in Rise of Skywalker where people kept saying she had, what, 20 or so seconds of screen time? Maybe it was 60, can't remember. Well, those people couldn't count, because I remember counting something like 147 seconds of screen time for her. That's over two minutes.

Is it still shittingly low for screen time? Absolutely. But it was still over 2x what the Twitter drama was going on about, and that was the point. A lot of these people today, don't even want to get their numbers straight, if it means they can gin up a bit of controversy.