Next-Gen PS5 & XSX |OT| Console tEch threaD

Status
Not open for further replies.
What proof is there that Microsoft has better decompression algorithms and "smarter texture streaming"?

Absolutely none.

There is abundant proof that Sony's I/O solution is significantly faster.

Yes the PS5 SSD is significantly faster, but the numbers we have now point to a decomp ratio of 2.0 on the XSX(2.4-4.8)and Cerny used 1.5 in the presentation(~7.5GB/s). That's why it's 8-9GB/s and not say 11GB/s for the PS5.

For the texture streaming the go read about sampler feedback streaming on the XSX or read the tweets from the developers.
 
Problem for whom? You?

I doubt MS is sticking a SSD demonstration in the middle of their third-party game showcase. They haven't done a proper hardware presentation yet, there is a time and place for that.
For me? No, man. For Houston :messenger_tears_of_joy:

If after all that instantaneous loading time mantra MS does not show how long the new gen games take to be fully loaded on the Series X, I will think that it is BS, yes.

But you know, that would be my personal opinion. You have another one? 🤷‍♂️
 
Last edited:
Yes the PS5 SSD is significantly faster, but the numbers we have now point to a decomp ratio of 2.0 on the XSX(2.4-4.8)and Cerny used 1.5 in the presentation(~7.5GB/s). That's why it's 8-9GB/s and not say 11GB/s for the PS5.

For the texture streaming the go read about sampler feedback streaming on the XSX or read the tweets from the developers.

Cerny said that the high end for PS5 was 22 GB/s in ideal situations.

You cannot take the numbers thrown out by MS and Sony and try and find the exact average. It's never going to be a perfect, steady number. It's going to depend on a ton of in-game situations.

The only thing we DO know is the stated uncompressed figures. Trying to rely on the 8-9 all the way up to 22 and comparing it to 4.8 is really not a valid comparison at all and doesn't say anything about the texture compression solution.

I'm also pretty sure there was a dev on REE that shot the rumor of some huge compression advantage down.
 
I'm also pretty sure there was a dev on REE that shot the rumor of some huge compression advantage down.

Honestly it's the final results that matter and not how the companies get there. It doesn't matter if one does compression better than the other if the final results are not superior to the competition.

PS5: Raw 5.5 GB/s Compressed 8-9 GB/s

Xbox Series X: Raw 2.4 GB/s Compressed 4.8 GB/s

Numbers from the manufacturers.
 
Honestly it's the final results that matter and not how the companies get there. It doesn't matter if one does compression better than the other if the final results are not superior to the competition.

PS5: Raw 5.5 GB/s Compressed 8-9 GB/s

Xbox Series X: Raw 2.4 GB/s Compressed 4.8 GB/s

Numbers from the manufacturers.
I agree the end number is which care, just please stop the specs are from each company not because you want the number will increase in some moment.

We cannot desistimate an Official 'data sheet' because my company has to be the best in all the things, yes the compresion of Xbox is better but only
for a part of the data and its raw speed is less the than the half of the other.

In a perfect world we should have the raw bandwidth of PS5, the compression for textures of XSX, the compression for other files of PS5, and IO of PS5 but
as this only fantasy this will not happens.

This are not theories only arguments without logical basis under the same "logical" you follow where XSX still working in improve their software
nobody stop PS5 of do the same so that argument in the best circumstance you will fight against an IO which do 22 GB/s.

What do you think when a dev excited about an SSD put as first example the PS5 ? is almost the double the data please stop before be start with an IO
hidden in the power supply.
 
Last edited:
I agree the end number is which care, just please stop the specs are from each company not because you want the number will increase in some moment.

We cannot desistimate an Official 'data sheet' because my company has to be the best in all the things, yes the compresion of Xbox is better but only
for a part of the data and his raw speed is less the than the half of the other.

In a perfect world we should have the raw bandwidth of PS5, the compression for textures of XSX, the compression for other files of PS5, and IO of PS5 but
as this only fantasy this will not happens.

This are not theories only arguments without logical basis under the same "logical" you follow where XSX still working in improve their software
nobody stop PS5 of do the same so that argument in the best circumstance you will fight against an IO which do 22 GB/s.

What do you think when a dev excited about an SSD put as first example the PS5 ? is almost the double the data please stop before be start with an IO
hidden in the power supply.

It's like if someone argues that the PS5 GPU is better because it's clock speed is higher. It's the results that matter more than anything and despite the PS5s higher clockspeed, it's GPU isn't as capable as the XSXs.
 
It's like if someone argues that the PS5 GPU is better because it's clock speed is higher. It's the results that matter more than anything and despite the PS5s higher clockspeed, it's GPU isn't as capable as the XSXs.

On paper, yes, that's true. The difference is PS5 has 15% less TFlops. However, with the additional clocks and GPU scrubbers and better I/O, it means that the GPU will punch above its weight and the reality will be less than 15% for probably the average case (there will be outliers in favor of a greater or less divide, some may even have PS5 games run at a higher metric of framerate or resolution).

But GPU isn't solely responsible for what you see on screen, so other factors do matter when it comes to texture detail and richness (such as the SSD speeds).

So yes, let's "SEE" what the "RESULTS" actually are. I think some that are expecting a universal across the board better image on the XSX are going to be in for a shocking awakening when comparisons or heavily optimized PS5 games are announced.
 
Cerny said that the high end for PS5 was 22 GB/s in ideal situations.

You cannot take the numbers thrown out by MS and Sony and try and find the exact average. It's never going to be a perfect, steady number. It's going to depend on a ton of in-game situations.

The only thing we DO know is the stated uncompressed figures. Trying to rely on the 8-9 all the way up to 22 and comparing it to 4.8 is really not a valid comparison at all and doesn't say anything about the texture compression solution.

I'm also pretty sure there was a dev on REE that shot the rumor of some huge compression advantage down.

I don't think it's fair to mention the 22GB/s max figure i.e a compression ratio of about 4:1 when MSFT hasn't mentioned any such figure.

But let me ask you though, when you calculate for 2.4 to 4.8 and 5.5 to 9, which one has a higher ratio? If you get the answer then maybe you'll come to the conclusion that one system as of what we know now, has a higher decomp ratio. It's just stating facts, one system is significantly faster the other seems to have better decompression as of now.
 
For me? No, man. For Houston :messenger_tears_of_joy:

If after all that instantaneous loading time mantra MS does not show how long the new gen games take to be fully loaded on the Series X, I will think that it is BS, yes.

But you know, that would be my personal opinion. You have another one? 🤷‍♂️
Sony hasn't shown anything either since the spiderman demo. And that's for a good reason; the actual performance is going to be unbelievable to consumers. MSFT made a mistake of showing those load times.
 
Honestly it's the final results that matter and not how the companies get there. It doesn't matter if one does compression better than the other if the final results are not superior to the competition.

PS5: Raw 5.5 GB/s Compressed 8-9 GB/s

Xbox Series X: Raw 2.4 GB/s Compressed 4.8 GB/s

Numbers from the manufacturers.
True that's why it's known the PS5 will have a much faster SSD.
 
I don't think it's fair to mention the 22GB/s max figure i.e a compression ratio of about 4:1 when MSFT hasn't mentioned any such figure.

But let me ask you though, when you calculate for 2.4 to 4.8 and 5.5 to 9, which one has a higher ratio? If you get the answer then maybe you'll come to the conclusion that one system as of what we know now, has a higher decomp ratio. It's just stating facts, one system is significantly faster the other seems to have better decompression as of now.
Better compression for streaming textures, yes. Although I still think a lot of this is apples to oranges comparison. Both SSD's are designed and built to do different things.
 
First of all Coreteks 100% knows his stuff and this was the first video i found wanting in some regards.

If we focus on my argument about eliminating bottlenecks, then Coreteks doesn't make a compelling argument(he doesn't for RT as well) based off what we now know. He mentions the custom controller in the PS5 and 12 channels as an advantage over the XSX, but this explains how the PS5 is able to have an impressive raw SSD speed of 5.5GB/s(>double the XSX) and not how it eliminates the IO bottlenecks between what's stored in secondary memory and the actual game code.

...But before continuing, I need to point out something we need to acknowledge that is misleading in Coretek's video. SSDs by default have low latency relative to HDDs, so both systems are going to change how RAM is utilized, the PS5's advantage here is the higher throughput...

In order to eliminate those bottlenecks, one can either rely solely on hardware accelerators or software or a combination of both. The data still has to go through the APU before going to RAM, so software on the CPU can be used to eliminate bottlenecks. Sony used a lot of custom hardware accelerators in the APU i.e they rely much on hardware, in order to spare the CPU. On the other hand, MSFT used a combination of Software and Hardware accelerators. The issue is, we do not know enough about MSFT's IO system, but what we do know so far is impressive and on par with the PS5's in terms of eliminating bottlenecks. MSFT is claiming that they only have File IO overhead equivalent to one tenth of a zen 2 core.

And my issue with Coreteks' video is that although he claims to be privy to some extra information, he's unable to clearly articulate(without revealing) what bottleneck the XSX may have. Instead he only mentions the public information from Cerny and infers that the XSX won't be able to eliminate these bottlenecks(which we now know is not true). The only thing he mentions about the XSX that was not publicly available at the time is that it has WDP compression but he doesn't mention BCPACK(which is the most important one) which at the time of his video was not publicly known to be present on the XSX.

Of the five bottlenecks presented by Cerny, the most important ones are File IO(including mapping) and Decompression. From what we know as of today, both systems eliminate these. For the others we'll find out but for example DirectStorage should be able to address DMA since it will be doing it for virtual RAM. So we still have a lot to learn.

I do not think we can talk about API or any software related stuff as that is impossible to tell for example I read somewhere PS5 SSD has 6 priority level compared to 2. As far as custom i/o is concerned I do not think there can be any doubt based on the diagram Sony showed in their presentation. That is definitely the best part about PS5 and that is why I think they spent more time than it
 
Xbox Streaming Textures / Velocity Architecture goal is 4.8gb/s constantly decompress, and will not go over this.

PS5 uses Kraken and achieves 8 or 9gb/s. Nothing more to say, the software will not make miracles and matches PS5's speed. If MS have a system to beat this, they were already saying for everyone just like the 12Tflops.

XSX is faster too, but PS5 is more.
The PS5 is powerful, but XSX is more.

Both great consoles, so this is it.
 
Last edited:
"Cross-gen" games are not going to get an actual "next-gen" version, these games are just compatible with the new machines. So they will have an enhanced mode when they run on the new machines (better frame rate/resolution/lighting effects/textures) in a similar way to what the PS4 PRO and Xbox one X offer right now... they are last gen games with a new coat of paint, devs can put as much or as little effort in it as they care for.

Case in point Gears 5, it looks better (better textures, lighting effects, reflections) but it has the same core, it doesn't offer anything that count not be reasonably downscaled to base xbox one.

This is why I find the whole idea of "cross buy" a bit disingenuous, we have backward compatible consoles, the idea is that there could be a PS4/xbox one game that does not work on the new machines is kind of funny... What it does not mean is that playstation and xbox will handle it the same, especially for existing games, MS seems to put a lot of emphasis on making old games look better anyway they can.

I mean no shit not sure why you think this needs explained lol

That's still a different version of the game... likely at least taking advantage of some new APIs, some different assets, etc.
 
Last edited:
Xbox Streaming Textures / Velocity Architecture goal is 4.8gb/s constantly decompress, and will not go over this.

PS5 uses Kraken and achieves 8 or 9gb/s. Nothing more to say, the software will not make miracles and matches PS5's speed. If MS have a system to beat this, they're saying for everyone just like the 12Tflops.

XSX is faster too, but PS5 is more.
The PS5 is powerful, but XSX is more.


Both great consoles, so this is it.
Took the words out of my mouth. Series X can't have it both ways with GPU and SSD, the console compromised on SSD in favour of a more powerful GPU (relatively speaking) and PS5 compromised on more powerful GPU in favour of a faster SSD (again relatively speaking.)

EDIT: sorry about my spelling, it's super late in my country and my brain is half functioning 😂
 
Last edited by a moderator:
I don't think it's fair to mention the 22GB/s max figure i.e a compression ratio of about 4:1 when MSFT hasn't mentioned any such figure.

But let me ask you though, when you calculate for 2.4 to 4.8 and 5.5 to 9, which one has a higher ratio? If you get the answer then maybe you'll come to the conclusion that one system as of what we know now, has a higher decomp ratio. It's just stating facts, one system is significantly faster the other seems to have better decompression as of now.

Or, maybe Microsoft just exaggerated their typical case and Sony understated it a little bit? We don't know. And it would be useful to know Microsoft's max figure. As someone mentioned. Is the top speed only 4.8 GB/s? That means the difference is even larger in favor of Sony's solution.

At the end of the day I am going with the base uncompressed speeds as being a good representation of the performance delta, and that compression techniques for each console won't amount to a meaningful difference.
 
Last edited:
I do not think we can talk about API or any software related stuff as that is impossible to tell for example I read somewhere PS5 SSD has 6 priority level compared to 2. As far as custom i/o is concerned I do not think there can be any doubt based on the diagram Sony showed in their presentation. That is definitely the best part about PS5 and that is why I think they spent more time than it
The 6 priority levels in the controller are definitely an advantage over competing systems but that's a whole different argument. SSDs have highly custom controllers and the XSX SSD controller should have some features not present in the PS5's.

The biggest difference is going to come from the higher throughput, on the PS5 and not any of these other things. But eliminating bottlenecks between the SSD and game code is more important than say the 6 priority levels. It doesn't matter how fast your SSD is, or how custom the controller is, if it can't load and stream assets faster into RAM.
 

That's the same thing as saying that the PS5s has secret SSD features that we don't know about. You really can't speculate on that kind of stuff without having an idea on what's in there. Otherwise you can just invent something crazy like a super secret high speed lane that will give instant access to all the data on the SSD.

In other words it's kind of pointless to discuss this type of thing because it's the same kind of thing that Mr X Media did.
 
Last edited:
That's the same thing as saying that the PS5s has secret SSD features that we don't know about. You really can't speculate on that kind of stuff without having an idea on what's in there. Otherwise you can just invent something crazy like a super secret high speed lane that will give instant access to all the data on the SSD.

In other words it's kind of pointless to discuss this type of thing because it's the same kind of thing that Mr X Media did.
I see you decided to miss my point.
 
What proof is there that Microsoft has better decompression algorithms and "smarter texture streaming"?

Absolutely none.

There is abundant proof that Sony's I/O solution is significantly faster.
so Sony's I/O is based on hardware and m$ on software? i am understanding right? i am talking about faster SSD and how it is archived?

because if that right, then m$ is behind and they will be even more and more later in PS5 gen..

yet, if i am not mistaken, Innocenceii said, that m$ is working years already on this software, if so, they can't really archived way more than they archived already, that is it, now they have just old friend mr. PR. consoles are in productions and soon they going to assemble the hardware.
 
The consoles really could use machine learnt upscale like Nvidia with DLSS.

MS have laid their cards on the table from what we can tell, but with Sony purported to have a few more things under their sleeve I wonder if there could be some solution yet to explore, especially with their VR aspirations.

Certainly AMD would be aiming to have a competitive solution next to Nvidia sooner rather than later. Why not save costs in developing the tech by partnering with a games company.

Granted, Sony studios have shown their prowess with generic upscaling like checkboarding as seen with Horizon, Spider-Man, etc. So aspirations over and above may not be a focus.

Thoughts inspired by this video....

 
Last edited:
Certainly AMD would be aiming to have a competitive solution next to Nvidia sooner rather than later. Why not save costs in developing the tech by partnering with a games company.

If I were to make a guess at something it would be a more advanced form of checkerboard rendering that's AI assisted.
 
Please stop this nonsense of the SSD(the compression and the demo thing) , this one of the few things
we have number you can fight for other thing... I don't know which IP for each company will have the first
two years, yes that one I least is more funny and you can make guess without bases.
 
Please stop this nonsense of the SSD(the compression and the demo thing) , this one of the few things
we have number you can fight for other thing... I don't know which IP for each company will have the first
two years, yes that one I least is more funny and you can make guess without bases.


Hopefully tomorrow brings us new topics to discuss.
 
SSD optimizations won't upgrade graphics of old games.

Of course not. I was referring to how the loading situation would look. I can see why it could be misinterpreted.

How much of an improvement would it be though? That's the real question.

Indeed. It will be interesting to see the first "real" examples, but I do suspect load times will look more like the fast resume feature they demoed.
 
but I do suspect load times will look more like the fast resume feature they demoed.

If it's loading from scratch they will be alot longer than the fast resume feature. Remember they are loading a saved state if the game which isn't the same as loading something from zero. Plus I'm guessing there's a limit to the amount of saved states that they can have.

Just my guess.
 
If it's loading from scratch they will be alot longer than the fast resume feature. Remember they are loading a saved state if the game which isn't the same as loading something from zero. Plus I'm guessing there's a limit to the amount of saved states that they can have.

Just my guess.

I think that when games are optimized for SSDs (and thus, might actually be smaller due to not needing duplication of textures), it will load a lot faster than we're used to - about the speed that we see in fast resume.

Time will tell :-)
 
SSD optimizations won't upgrade graphics of old games.
It could assist in with graphical issues such as texture and LOD pop in's in certain titles, even the occasional save or load stutter of some titles. However there aren't many games that suffer from such matters this gen that I can think off.
 
"Optimized for Series X" is the new "Xbox One X Enhanced". Lockhart confirmed?

If Lockhart is "Series S" (or another letter), what series is it then if both of them have one console? That branding is a big mess. It also clearly shows how Microsoft doesn't give a single fuck about people whose first language isn't English because the name is very difficult to pronounce. I see people just calling it "Xbox" and ending up with XSX at home and a lot of surprises.
 
It could assist in with graphical issues such as texture and LOD pop in's in certain titles, even the occasional save or load stutter of some titles. However there aren't many games that suffer from such matters this gen that I can think off.

Well, DOOM 2016 and all other Mega
Texture based games do have quite an issue with texture pop in. Not sure how much can that be improved without a patch though.
 
Status
Not open for further replies.
Top Bottom