• 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.
  • The Politics forum has been nuked. Please do not bring political discussion to the rest of the site, or you will be removed. Thanks.

(*)SSD and loading times demystified

Oct 23, 2016
410
414
500
28
Argentina
You can. It's called "hypervisor".
If your OS runs as yet another software under the same hypervisor.
And hypervisors can have pretty small footprint in memory.
But I've already mentioned that it depends how many services you want to run in background.
If "everything" then obviously the whole OS runs all the time.
Yeah, but hypervisors are used for OSes, not programs. You cannot put a program under the control of a hypervisor. Hypervisor let many OSes work on the same hardware at the "same" time, but each program has to run under their respective OS.

That's a too literal understanding of what I mean.
Well, you didn't explain much
 

psorcerer

Banned
May 1, 2012
2,479
3,537
1,035
How many games have you programed in you life?

Zero. Actually.
If I was under NDA I wouldn't speak so freely about things.
But I do know a thing or two about hardware and low level optimizations: GPU machine learning, high perf computing, distributed computations.
 

Shmunter

Member
Aug 25, 2018
10,764
24,390
835
Sounds good. I think that closes the gap compared to the PS5 SSD.
XsX can do exactly the same shit as PS5 with the SSD, just nowhere near as fast or snappy.

It’s still a massive leap for pushing this tech forward that will be amazing.
 

psorcerer

Banned
May 1, 2012
2,479
3,537
1,035
Yeah, but hypervisors are used for OSes, not programs. You cannot put a program under the control of a hypervisor. Hypervisor let many OSes work on the same hardware at the "same" time, but each program has to run under their respective OS.

From the hypervisor perspective any OS is a "program" and there is not difference between real "program" and "OS" one.
All these razor thin definitions are useless in custom hardware where you don't need to maintain existing status quo.
If you look at a modern hypervisor like KVM you can see that it treats each OS as a mere process in its address space. There is not real difference between user space process and OS to them.
It all depends on what services need to be provided to the "program". OS needs less services than the real user-space software.
 

Shmunter

Member
Aug 25, 2018
10,764
24,390
835
This really clears many things up.
From what I understand the CPU knows what textures/assets need loading depending on player proximity and uses semaphores to know when to check for new data on the RAM and then inform the GPU to render that asset.
Also semaphores have absolutely nothing to do with what assets currently need loading and is just a command that gets executed every time the GPU ends a frame(for eg) to notify the CPU, and thus are able to render whatever the CPU has loaded onto the RAM from SSD, correct? Is this the solution you expect most games to employ going forward?

Also instead of proximity LOD, in the scenario where the player is turning his camera and new assets need to load in as he turns, the same technique can be used right? For proximity, its sensible that there are just a few assets loading in as the player moves, but when you turn the camera there's suddenly a whole truckload of new assets that you need to stream in. I watched the Horizon video and Cyberpunk 2077 takes this to even more of an extreme where its loading in new assets not just for horizontal camera rotation , but vertical rotation too. So what you see when you turn the camera upwards is entirely new content just loaded in discarding all the data that was present when you were looking straight. Do you think both these games use the solution mentioned above?
Also when there needs to be a huge number of assets that need streaming in one frame, how does it affect performance? I'm assuming it doesn't affect performance at all.

Also I'd like a more in depth explanation of the GPU cache scrubber present in the PS5. This seems to be the main key in its arsenal, and what most people here seem to be missing when they say things like ''but the xbox has SSD too''. Clearing this thing up could help with lot of the issues here I think.
Great thought process.

The techniques you mention and many others are likely to be used in streaming. The limits have been raised so dramatically all the stuff you mentioned will be almost standard in many titles. Can you imagine?

Regarding tech stuff, there should be no performance penalty in streaming, the i/o chips handle it independent of cpu\gpu and stream into ram where the cpu/GPU picks it up like nothing happened. Programmers will tune it for their particular need too. Just like they have with current gen games e.g. Spider-Man no hitching ever through an open city. But now 100x more power and potential.

I suspect the GPU/cpu can also directly access the ssd data, so they may be able to fetch some things without going to ram. No idea what, but it’s possible I think.

The scene detail stuff nextgen is what I’m really mostly excited to see.

Both next gen system will do things like this.
 
Last edited:

KlickKlack

Banned
Dec 26, 2018
494
619
335
Data per frame is much more important than fps and resolutions, they are both 4k machines anyway. Playstations ssd is awesome and we all expected it but fucking hell all the complications happened because of the stupid low 16gb and god knows itll be a bottleneck my preferred pool was 24-32gb atleast not 16
Sony says official that all the ssd tech has one big function: reduction of the loading times:

PlayStation. blog

"As Mark discussed in his presentation, PS5’s ultra-high-speed SSD and integrated custom I/O system were developed with the goal of removing barriers to play – specifically loading screens. Developers are able to stream assets into PS5 games at an incredibly fast rate, so PS5 play experiences can be seamless and dynamic, with near-instantaneous fast travel through large game worlds. This enhanced speed will enable game developers to create larger, richer worlds without traditional limitations, such as load times, and also allows gamers to spend more time gaming than waiting."


Nothing about "magical power increase" because of the SSD.
 
Last edited:

hemo memo

Member
Sep 13, 2011
8,763
7,717
1,045
Certain people are downplaying the impact of SSD is there a reason? I thought Cerny talk about that was clear yet some people are still confused.
 
  • Like
Reactions: Mokus

hyperbertha

Member
Nov 24, 2018
2,225
7,283
655
Great thought process.

The techniques you mention and many others are likely to be used in streaming. The limits have been raised so dramatically all the stuff you mentioned will be almost standard in many titles. Can you imagine?

Regarding tech stuff, there should be no performance penalty in streaming, the i/o chips handle it independent of cpu\gpu and puts it into ram where the cpu/GPU picks it up like nothing happened. Programmers will tune it for their particular need too. Just like they have with current gen games e.g. Spider-Man no hitching ever through an open city. But now 100x more power and potential.

The scene detail stuff increases are what I’m really mostly excited to see.
Yea its going to be big.
Unless I'm mistaken, Spiderman achieves it on ps4 by asset repetition, ie instancing the same models, NPCs and textures multiple times as you zip through the city so they don't fill up memory space. With SSDs you no longer have that problem. Its an exponential increase in variety, not just the quality or details of assets
 
Last edited:

vpance

Member
Jun 20, 2005
9,126
3,451
1,650
So the decompressor is apparently equivalent to 9 Zen 2 cores. That seems really impressive. Or is decompression something easily optimized for on custom logic? I wonder how much die space it takes compared to the CPU.

So that lets them achieve a possible 22GB/s on decompression. BG a dev on here has said the average transfer rate he saw was like 20GB/s. That's insane if true.
 
  • Fire
Reactions: megreotsugua

psorcerer

Banned
May 1, 2012
2,479
3,537
1,035
Or is decompression something easily optimized for on custom logic? I wonder how much die space it takes compared to the CPU.

Probably a very small fixed function logic.
So that lets them achieve a possible 22GB/s on decompression.

Essentially it's the maximum throughput of the I/O stack.
It will probably rarely be achieved in a real software though.
 
Oct 23, 2016
410
414
500
28
Argentina
From the hypervisor perspective any OS is a "program" and there is not difference between real "program" and "OS" one.
All these razor thin definitions are useless in custom hardware where you don't need to maintain existing status quo.
If you look at a modern hypervisor like KVM you can see that it treats each OS as a mere process in its address space. There is not real difference between user space process and OS to them.
It all depends on what services need to be provided to the "program". OS needs less services than the real user-space software.
Yes, I know the hypervisor will treat any OS as if it were a program. What I'm saying is that you can't put a program (a game, which runs under an Operative System) in that space because the hypervisor wouldn't understand it (and vice versa).
If you say that the hypervisor has the services to understand that program, then the hypervisor would be the OS you are trying to replace, which makes all this redundant.

I agree with what Shmunter Shmunter said. You could release the GUI to free resources, but not the whole OS.
 
Dec 1, 2018
1,520
1,240
530
Sony says official that all the ssd tech has one big function: reduction of the loading times:

PlayStation. blog

"As Mark discussed in his presentation, PS5’s ultra-high-speed SSD and integrated custom I/O system were developed with the goal of removing barriers to play – specifically loading screens. Developers are able to stream assets into PS5 games at an incredibly fast rate, so PS5 play experiences can be seamless and dynamic, with near-instantaneous fast travel through large game worlds. This enhanced speed will enable game developers to create larger, richer worlds without traditional limitations, such as load times, and also allows gamers to spend more time gaming than waiting."


Nothing about "magical power increase" because of the SSD.
Cerny said alot especially in his ps5 developer deepdown video, it is u and ur tiny brain that doesnt seem to understand it.
 

demigod

Member
Apr 17, 2007
8,708
7,974
1,605
I can’t wait for next gen. SSD will be huge. This will be a major boom to games that require constant loading.

Before Bandai patched DBZ: Kakarot, it took about 30secs to load between zones. Its about half that now. Imagine loading 5-7 zones to collect the dragon balls and repeating it over and over. This is on ps4 pro with ssd.

I’m currently playing monster hunter iceborne on my 2070 laptop with an ssd. It takes me 16 secs(saw a ps4 video taking 30secs with ssd) to load into Guiding Lands and about another 5 sec to load back to town. It might not look long but i’m constantly having to load over and over to get the mob I want. Something that takes me 5-10mins will probably take me about 1 minute next gen.
 
  • Fire
Reactions: megreotsugua

thelastword

Banned
Apr 7, 2006
11,732
11,949
2,000
@VFXVeteran, what do you think about this thread?
Ha ha, good luck with that...VFX will just downplay it a bit more.....

Not surprised this thread didn't get the same amount of replies as the other thread since this is based on real-life technical evidence.
Hey, this thread was buried.....I never saw it till someone else pasted the OP's post in another thread......Thankfully he did, nice breakdown.....I guess not much discussion ensued because it's not a "My 12TF is greater than your TF Dawg, end discussion"....
 

vdopey

Gold Member
Jun 13, 2019
179
308
410
Mm you can't do that. When you suspend, you are only powering off everything except the RAM.
Also, you can't remove the OS from RAM, everything else stops working. You always need the OS in RAM because the OS it's the system administrator. Any program (games in this case) that's launched, has to ask for resources or other things to the OS

I've mentioned this here: https://www.neogaf.com/threads/next...-analysis-leaks-thread.1480978/post-257471394

Essentially the kernel will always be resident, this manages the hardware and if anything like Linux the Freebsd Kernel will use around 40 - 80MB, it will also handle the virtualisation, the PS5 "OS" or the user interface sits inside a virtualized container, this can be suspended and resumed on the fly every time you request from the ps5 controller within a fraction of a second it can resume the os and then suspend back to the game. Also the read in will require the full size to be read in to the RAM (for obvious reasons), but the suspend back will only need to be the delta of whats changed in the state between the last time until now. The ssd is stateful / permanent, so it doesn't need the full 2.5G or whatever each time just the delta change to be written back to the ssd storage so it might be as little as 100M back to ssd to resume, it may even be 0 written back if nothing has changed.

The tricky part becomes memory resident applications, for example spotify or voice chat / party chat or worst case game sharing or even downloading while playing. So this is why I thought maybe the whole lot will be memory resident, but further thought on it, why would it need to be ? Every Game built on the PS5 uses the PS5 apis, so the game developers if they want to allow music to be played while the game is running will have to enable that within their game, which means they may have to deal with the possible fallout of spotify running in the background consuming cpu and ram. So some sort of api listener will always need to be memory resident, or does it ? that listener wont be consuming much ram though its just that it will also possibly have to reserve ram for game sharing or voice chat and possibly music - I reckon at max probably 1G reserved, maybe as low as 500M none of these services are hugely memory hungry other than possibly game sharing - the actual Playstation User Interface would be the most resource hungry, but that doesn't need to run and if they can offload that memory utilisation to the api and therefore the gamedevs themselves it means they will only need to reserve a tiny amount and let the game devs and the game itself manage how much ram those other services can take up / use. I wouldn't be surprised if 15G is available to Game Devs and 1G reserved, it may even be as high as 15.5G, if this is the strategy they are going with and probably 1 thread in the 16 thread cpu. Again they might be using a custom ARM processor for the OS and completely offloading all of that to that, lets not forget the dedicated hardware for audio, so things like spotify / voice chat may all use that and not even hit the actual cpu.

Sony do need to clarify how much of the custom silicon from the PS4 is being carried across, as they had custom ARM based silicon for the sharing button. I expect in the next few weeks / month a lot more will be revealed.

I don't know if their strategy will pay off, or even if this is their strategy. It all comes down to cost at this point that SSD is damn expensive, how much are they paying for it ?

Their R&D budget must have also been high for all of the custom silicon like the gpu scrubbers io controller, audio etc, how much have they patented ? they wont need to pay royalties for any of that back to AMD but they may be able to get royalties from the patents if their designs work. Cerny did mention design aspects of their custom silicon may be seen in current and future amd silicon, so amd might have done an agreement with sony to not charge them royalties for their new designs for access to sonys custom stuff. I think I read somewhere Sony has a bunch of patents around RayTracing as well. I reckon Microsoft will be paying a lot more to AMD per console than Sony, this will all remain to be seen.
 

MetalAlien

Banned
Mar 6, 2005
13,609
11,379
1,940
If one or the other does have a clear advantage only first party or exclusive games will show it. Everyone else is going to write to the lowest common denominator.
 
  • Like
Reactions: Three

Zekra Dezivad

Member
Aug 22, 2011
1,662
735
880
If one or the other does have a clear advantage only first party or exclusive games will show it. Everyone else is going to write to the lowest common denominator.
This is a bit generic. If within the same game you can have advantages you'll probably have them.
Better hardware can have better resolution, at the same time a faster configuration can have faster loading times. This is just an example, but I think it's very clear. I suppose 2/3TF of difference or the sdd speed difference won't be enough to have drammatically different games across the systems.
 

MetalAlien

Banned
Mar 6, 2005
13,609
11,379
1,940
This is a bit generic. If within the same game you can have advantages you'll probably have them.
Better hardware can have better resolution, at the same time a faster configuration can have faster loading times. This is just an example, but I think it's very clear. I suppose 2/3TF of difference or the sdd speed difference won't be enough to have drammatically different games across the systems.
Well i'm just saying if one of the consoles can take advantage of I dunno, Sony's very fast SSD and so something the Xbox can't... or even the other way around... they probably won't to get the games on both consoles looking the same.
 

RayHell

Member
Jan 11, 2020
333
634
320
This will be good for 1st party but what about third party that need to support multiple platform with slower io. They still need those long corridor for assets loading or even memory pool chunk streaming. Do you think we’ll still see a gain other than loading/fast travel time?
 

psorcerer

Banned
May 1, 2012
2,479
3,537
1,035
This will be good for 1st party but what about third party that need to support multiple platform with slower io. They still need those long corridor for assets loading or even memory pool chunk streaming. Do you think we’ll still see a gain other than loading/fast travel time?

It depends. If games will start at least requiring SSDs then it will move.
Otherwise it won't change a thing.
 

vdopey

Gold Member
Jun 13, 2019
179
308
410
So in those cases would you say PS5 have a disadvantage over XSX?

I dont think anyone has dismissed the raw power of the xsx side, it is more powerful than the PS5, there is no denying that - no trickery, no magic. Microsoft has produced a beast, but also it has no mystery around, the design is not too dissimilar to what you would find in a good gaming pc, except in a tiny form factor and its designed really well - I really like how it all slots together and I like the tower design, its like a mini pc but packs a big punch.

The reason why there are so many PS threads and so much speculation around the PS5, is because Sony have taken a different and imho much braver route, they haven't gone for raw power but instead are trying to improve / squeeze out as much performance and are using parts not seen or used in the PC space yet. Once the initial shock / disappointment wears off and you start looking past the raw power disadvantage and looking at what they are proposing it becomes really interesting. It may end up producing a completely different approach to game design, or it may become a complete dud, its all dependant on how difficult it is to optimise for and if its worth the time, which will largely depend on how many units Sony sell, but also Microsoft has tried to go down a similar route, its a lot more standard pc based, but its inclusion does give 3rd parties an onus to at least attempt to change game design. I don't know the impact on PC gaming, specs for modern open world games on pcs may now include a requirement for SSD's of a specific capability, who knows. Sonys first party will by and large be the ones to take great advantage of this - lets see what HZD2 looks like on the PS5, Guerrilla Games used KillZone to showcase the PS4 and no doubt they will be using HZD2 to showcase what all of these various optimised silicon components Sony have conjured up allows the PS5 to do.

Its ultimately just fun, a lot of us are locked at home due to this COVID-19, I have never commented on here before I think I started this weekend, I am really quite bored trying to stop myself going out, so this has become a release and its fun to speculate, there are aspects of Unix / Linux & Software I really understand well, Psorcerer and some others here also really clued up and its fun to discuss this stuff without flinging shit at each other. I may well end up getting an xsx, but I really really really want to put Linux on it, its a great form factor and would make a great steam machine :p
 
Last edited:

RayHell

Member
Jan 11, 2020
333
634
320
I dont think anyone has dismissed the raw power of the xsx side, it is more powerful than the PS5, there is no denying that - no trickery, no magic. Microsoft has produced a beast, but also it has no mystery around, the design is not too dissimilar to what you would find in a good gaming pc, except in a tiny form factor and its designed really well - I really like how it all slots together and I like the tower design, its like a mini pc but packs a big punch.

The reason why there are so many PS threads and so much speculation around the PS5, is because Sony have taken a different and imho much braver route, they haven't gone for raw power but instead are trying to improve / squeeze out as much performance and are using parts not seen or used in the PC space yet. Once the initial shock / disappointment wears off and you start looking past the raw power disadvantage and looking at what they are proposing it becomes really interesting. It may end up producing a completely different approach to game design, or it may become a complete dud, its all dependant on how difficult it is to optimise for and if its worth the time, which will largely depend on how many units Sony sell, but also Microsoft has tried to go down a similar route, its a lot more standard pc based, but its inclusion does give 3rd parties an onus to at least attempt to change game design. I don't know the impact on PC gaming, specs for modern open world games on pcs may now include a requirement for SSD's of a specific capability, who knows. Sonys first party will by and large be the ones to take great advantage of this - lets see what HZD2 looks like on the PS5, Guerrilla Games used KillZone to showcase the PS4 and no doubt they will be using HZD2 to showcase what all of these various optimised silicon components Sony have conjured up allows the PS5 to do.

Its ultimately just fun, a lot of us are locked at home due to this COVID-19, I have never commented on here before I think I started this weekend, I am really quite bored trying to stop myself going out, so this has become a release and its fun to speculate, there are aspects of Unix / Linux & Software I really understand well, Psorcerer and some others here also really clued up and its fun to discuss this stuff without flinging shit at each other. I may well end up getting an xsx, but I really really really want to put Linux on it, its a great form factor and would make a great steam machine :p
I don’t want to dismiss anything, it’s a very informative thread and I want to make sure I understood every part of it. I was simply wondering what are the limits of those well engineered consoles. Like you said its pretty fun during those time and I’m so excited about next gen right now. I think it’s going to the first time since PS One that console will have an edge on PC with this kind of data steaming because its used in a closed ecosystem.
 
  • Like
Reactions: vdopey

Audiophile

Member
Jun 25, 2015
718
2,374
700
England, UK
Sony says official that all the ssd tech has one big function: reduction of the loading times:

PlayStation. blog

"As Mark discussed in his presentation, PS5’s ultra-high-speed SSD and integrated custom I/O system were developed with the goal of removing barriers to play – specifically loading screens. Developers are able to stream assets into PS5 games at an incredibly fast rate, so PS5 play experiences can be seamless and dynamic, with near-instantaneous fast travel through large game worlds. This enhanced speed will enable game developers to create larger, richer worlds without traditional limitations, such as load times, and also allows gamers to spend more time gaming than waiting."


Nothing about "magical power increase" because of the SSD.

Nobody is saying it causes a "magical power increase"...That's just being obtuse.

It's about removing an enormous bottleneck that has the ability to change the fundamental nature of game worlds; and gives developers the opportunity to "create larger, richer worlds without traditional limitations" as you've quoted above. You'll also be able to traverse and travel in time, space and scale faster than ever in these game worlds too.

When they say "..such as load time", they're simply providing the most straightforward and easy to understand example of where games stand to benefit. There's a lot more to it than "stuff loads quicker".
 

David Ricardo

Member
Sep 25, 2008
1,395
1,002
1,275
Gijon, Spain
There's a lot of confusion on why SSD is so important for next-gen and how it will change things.
Here I will try to explain the main concepts.
TL;DR fast SSD is a game changing feature, this generation will be fun to watch!

It was working fine before, why do we even need that?
No, it wasn't fine, it was a giant PITA for anything other than small multiplayer maps or fighting games.
Let's talk some numbers. Unfortunately not many games have ever published their RAM pools and asset pools to the public, but some did.
Enter Killzone: Shadowfall Demo presentation.
We have roughly the following:

TypeApprox. Size, %Approx. Size, MB
Textures30%1400
CPU working set15%700
GPU working set25%1200
Streaming pool10%500
Sounds10%450
Meshes10%450
Animations/Particles1%45

*These numbers are rounded sums of various much more detailed numbers presented in the article above.

We are interested in the "streaming pool" number here (but we will talk about others too)
We have ~500MB of data that is loaded as the demo progresses, on the fly.
The whole chunk of data that the game samples from (for that streaming process) is 1600MB.
The load speed of PS4 drive is (compressed data) <50MB/sec (uncompressed is <20MB/sec), i.e. it will take >30sec to load that at least.

It seems like it's not that big of a problem, and indeed for demo it is. But what about the game?
The game size is ~40GB, you have 6.5GB of usable RAM, you cannot load the whole game, even if you tried.
So what's left? We can either stream things in, or do a loading screen between each new section.
Let's try the easier approach: do a loading screen
We have 6.5GB of RAM, and the resident set is ~2GB from the table above (GPU + CPU working set). We need to load 4.5GB each time. It's 90 seconds, pretty annoying, but it's the best case. Any time you need to load things not sequentially, you will need to seek the drive and the time will increase.
You can't go back, as it will re-load things and - another loading screen.
You can't use more than 4.5GB assets in your whole gaming section, or you will need another loading screen.
It gets even more ridiculous if your levels are dynamic: left an item in previous zone? Load time will increase (item is not built into the gaming world, we load the world, then we seek for each item/item group on disk).
Remember Skyrim? Loading into each house? That's what will happen.
So, loading screens are easy, but if your game is not a linear, static, theme-park style attraction it gets ridiculous pretty fast.

How to we stream then?
We have a chunk of memory (remember 500Mb) that's reserved for streaming things from disk.
With our 50MB/sec speed we fill it up each 10 sec.
So, each 10 sec we can have a totally new data in RAM.
Let's do some metrics, for example: how much new shit we can show to the player in 1 min? Easy: 6*500 = 3GB
How much old shit player sees each minute? Easy again: 1400+450+450+45=~ 2.5GB
So we have a roughly 50/50 old to new shit on screen.
Reused monsters? assets? textures? NPCs? you name it. You have the 50/50 going on.

But PS4 has 6.5GB of RAM, we used only 4.5GB till now, what about other 2GB?
Excellent question!
The answer is: it goes to the old shit. Because if we increase the streaming buffer to 1.5GB it still does nothing to the 50MB/sec speed.
With the full 6.5GB we get to 6GB old vs 3GB new in 1 minute. Which is 2:1 old shit wins.

But what about 10 minutes?
Good, good. Here we go!
In 10 min we can get to 30GB new shit vs 6GB old.
And that's, my friends, how the games worked last gen.
You're as a player were introduced to the new gaming moments very gradually.
Or, there were some tricks they used: open doors animation.
Remember Uncharted with all the "let's open that heavy door for 15sec?" that's because new shit needs to load, players need to get to a new location, but we cannot load it fast.

So, what about SSDs then?
We will answer that later.
Let's ask something else.

What about 4K?
With 4K "GPU working set" will grow 4x, at least.
We are looking at 1200*4 = 4.8GB of GPU data.
CPU working set will also grow (everybody wants these better scripts and physics I presume?) but probably 2x only, to 700*2 = ~1.5GB
So overall the persistent memory will be well over 6GB, let's say 6.5GB.
That leaves us with ~5GB of free RAM in XSeX and ~8GB for PS5.

Stop, stop! Why PS5 has more RAM suddenly?
That's simple.
XSeX RAM is divided into two pools (logically, physically it's the same RAM): 10GB and 3.5GB.
GPU working set must use the 10GB pool (it's the memory set that absolutely needs the fast bandwidth).
So 10 - 4.8 = 5.2 which is ~5GB
CPU working set will use 3.5GB pool and we will have a spare 2GB there for other things.
We may load some low freq data there, like streaming meshes and stuff, but it will hard to use in each frame: accessing that data too frequently will lower the whole system bandwidth to 336Mb/sec.
That's why MSFT calls the 10GB pool "GPU optimal".

But what about PS5? It also has some RAM reserved for the system? It should be ~14GB usable!
Nope, sorry.
PS5 has a 5.5GB/sec flash drive. That typically loads 2GB in 0.27 sec. It's write speed is lower, but not less than 5.5GB/sec raw.
What PS5 can do, and I would be pretty surprised if Sony won't do it. Is to save the system image to the disk while the game is playing.
And thus give almost full 16GB of RAM to the game.
2GB system image will load into RAM in <1 sec (save 2GB game data to disk in 0.6 sec + load system from disk 0.3 sec). Why keep it resident?
But I'm on the safe side here. So it's ~14.5GB usable for PS5.

Hmm, essentially MSFT can do that too?
Yep, they can. The speeds will be less sexy but not more than ~3sec, I think.
Why don't they do it? Probably they rely on OS constantly running on the background for all the services it provides.
That's why I gave Sony 14.5GB.
But I have hard time understanding why 2.5GB is needed, all the background services can run on a much smaller RAM footprint just fine, and UI stuff can load on-demand.

Can we talk about SSD for games now?
Yup.
So, let's get to the numbers again.
For XSeX ~5GB of "free" RAM we can divide it into 2 parts: resident and streaming.
Why two? Because typically you cannot load shit into frame while frame is rendering.
GPU is so fast, that each time you ask GPU "what exact memory location are you reading now?" will slow it down to give you an answer.

But can you load things into other part while the first one is rendering?
Absolutely. You can switch "resident" and "streaming" part as much as you like, if it's fast enough.
Anyway, we got to 50/50 of "new shit" to "old shit" inside 1 second now!
2.5GB of resident + 2.5GB of streaming pool and it takes XSeX just 1 sec to completely reload the streaming part!
In 1 min we have 60:1 of new/old ratio!
Nice!

What about PS5 then? Is it just 2x faster and that's it?
Not really.
The whole 8GB of the RAM we have "free" can be a "streaming pool" on PS5.

But you said "we cannot load while frame is rendering"?
In XSeX, yes.
But in PS5 we have GPU cache scrubbers.
This is a piece of silicon inside the GPU that will reload our assets on the fly while GPU is rendering the frame.
It has full access to where and what GPU is reading right now (it's all in the GPU cache, hence "cache scrubber")
It will also never invalidate the whole cache (which can still lead to GPU "stall") but reload exactly the data that changed (I hope you've listened to that part of Cerny's talk very closely).

But it's free RAM size doesn't really matter, we still have 2:1 of old/new in one frame, because SSD is only 2x faster?
Yes, and no.
We do have only 2x faster rates (although the max rates are much higher for PS5: 22GB/sec vs 6GB/sec)
But the thing is, GPU can render from 8GB of game data. And XSeX - only from 2.5GB, do you remember that we cannot render from the "streaming" part while it loads?
So in any given scene, potentially, PS5 can have 2x to 3x more details/textures/assets than XSeX.
Yes, XSeX will render it faster, higher FPS or higher frame-buffer resolution (not both, perf difference is too low).
But the scene itself will be less detailed, have less artwork.

OMG, can MSFT do something about it?
Of course they will, and they do!
What are the XSeX advantages? More ALU power (FLOPS) more RT power, more CPU power.
What MSFT will do: rely heavily on this power advantage instead of the artwork: more procedural stuff, more ALU used for physics simulation (remember, RT and lighting is a physics simulation too, after all).
More compute and more complex shaders.

So what will be the end result?
It's pretty simple.
PS5: relies on more artwork and pushing more data through the system. Potentially 2x performance in that.
XSeX: relies more on in-frame calculations, procedural. Potentially 30% performance in that.
Who will win: dunno. There are pros and cons for each.
It will be a fun generation indeed. Much more fun than the previous one, for sure.

P.S.
adding the bonus rounds here:

But what the 60:1 new/old ratio will give us, gamers?
Did you ever wonder why none of the open world game ever featured a city?
Like real city, with blocks, apartments, etc.?
Because it would be boring as fuck.
You will enter each and every room to see almost exactly the same props and furniture.
Exactly the same NPCs. Doing same things.
With next get flash drive an apartment in the city can use the whole non-resident RAM space to render that.
And load a new one when you go from one door to another.
Think about it. 1sec 5.5Gb. Behind each door you can have a full new world. Completely different from the other.

But, but I swapped HDD to SSD in my PS4 and nothing changed! What's going on?
You see. Let's return to that Killzone example.
We have that 500MB streaming buffer and we load it with the new data in 10 sec, on HDD.
Now we swapped in SSD with 500MB/sec and we load that buffer in 1 sec!
But, guess what, game was not designed around that.
Game was designed to demand and use that data in next 10 sec, not right now.
So, no matter how fast your SSD is, it will not change anything.
I would say in a properly designed game it will even make things worse.
Why? Because you used 500MB/sec of memory bandwidth right now, for the data that won't be needed at all until game will require it 10 sec later, wasted 500Mb/sec for nothing instead of giving it to the GPU!

But how do I know if game was designed around SSD?
Simple. It will not work on HDD, like at all.
Like giving you 1fps, 0.2fps, complete slideshow.
Unless it behaves like that. It's not a game for SSD.
Amazing OP


I was initially dissappointed by PS5s specs, but your post gave me something to look forward to.
 
  • Like
Reactions: Entroyp and sircaw
Jun 16, 2018
4,302
3,946
550
prettycoolgraphics.blogspot.com
Also even though the ps5 can load much higher quality assets in 1 sec due to having 8 GB of streaming to work with every second, doesn't all that higher quality still put a load on the GPU since the frame will have more quality, hence more difficult to render, hence affecting frame rates? This kinda relates to my previous question as the answer depends on how it juggles between frames and streamed data that it gets every 1 sec.
The ps4 can handle extremely detailed assets as seen in many games. You aren't looking at all of them close up at once. Same applies for next gen. The problem is that it seems current games are only using a fraction of currently available memory which is likely part of the reason you still see blurry textures and many low priority objects at significantly less detail. For more detail see further down this comment.
Nothing about "magical power increase" because of the SSD.
I'll quote myself

Radeon RX 580 4GB Game FPS Benchmarks On Ultra


Here we see that at 1080p, the 4GB can run most games on ultra at decent to high framerate. Even at 1440p ultra it gets close to 30fps up to 60 fps, if I'm not mistaken.

Ultra settings are higher than anything on console, yet 4GB is enough to handle it currently. That suggests less than 4GB of graphics might be usable in any frame in current gen games.
Sure an rtx 2080 has significantly more ram and more Tflops, but it won't be able to show more details than ultra at 1080p or 1440p. In essence it can only increase framerate, but detail wise it's limited to what the 4GB rx 580 can display.(at least in current gen titles without ray tracing.). Why are games using so little ram? Could very well be due to lack of ssd. Would a game that actually used 8GB instead of 4GB not look better?
 
Last edited:
Oct 23, 2016
410
414
500
28
Argentina
I've mentioned this here: https://www.neogaf.com/threads/next...-analysis-leaks-thread.1480978/post-257471394

Essentially the kernel will always be resident, this manages the hardware and if anything like Linux the Freebsd Kernel will use around 40 - 80MB, it will also handle the virtualisation, the PS5 "OS" or the user interface sits inside a virtualized container, this can be suspended and resumed on the fly every time you request from the ps5 controller within a fraction of a second it can resume the os and then suspend back to the game. Also the read in will require the full size to be read in to the RAM (for obvious reasons), but the suspend back will only need to be the delta of whats changed in the state between the last time until now. The ssd is stateful / permanent, so it doesn't need the full 2.5G or whatever each time just the delta change to be written back to the ssd storage so it might be as little as 100M back to ssd to resume, it may even be 0 written back if nothing has changed.

The tricky part becomes memory resident applications, for example spotify or voice chat / party chat or worst case game sharing or even downloading while playing. So this is why I thought maybe the whole lot will be memory resident, but further thought on it, why would it need to be ? Every Game built on the PS5 uses the PS5 apis, so the game developers if they want to allow music to be played while the game is running will have to enable that within their game, which means they may have to deal with the possible fallout of spotify running in the background consuming cpu and ram. So some sort of api listener will always need to be memory resident, or does it ? that listener wont be consuming much ram though its just that it will also possibly have to reserve ram for game sharing or voice chat and possibly music - I reckon at max probably 1G reserved, maybe as low as 500M none of these services are hugely memory hungry other than possibly game sharing - the actual Playstation User Interface would be the most resource hungry, but that doesn't need to run and if they can offload that memory utilisation to the api and therefore the gamedevs themselves it means they will only need to reserve a tiny amount and let the game devs and the game itself manage how much ram those other services can take up / use. I wouldn't be surprised if 15G is available to Game Devs and 1G reserved, it may even be as high as 15.5G, if this is the strategy they are going with and probably 1 thread in the 16 thread cpu. Again they might be using a custom ARM processor for the OS and completely offloading all of that to that, lets not forget the dedicated hardware for audio, so things like spotify / voice chat may all use that and not even hit the actual cpu.

Sony do need to clarify how much of the custom silicon from the PS4 is being carried across, as they had custom ARM based silicon for the sharing button. I expect in the next few weeks / month a lot more will be revealed.

I don't know if their strategy will pay off, or even if this is their strategy. It all comes down to cost at this point that SSD is damn expensive, how much are they paying for it ?

Their R&D budget must have also been high for all of the custom silicon like the gpu scrubbers io controller, audio etc, how much have they patented ? they wont need to pay royalties for any of that back to AMD but they may be able to get royalties from the patents if their designs work. Cerny did mention design aspects of their custom silicon may be seen in current and future amd silicon, so amd might have done an agreement with sony to not charge them royalties for their new designs for access to sonys custom stuff. I think I read somewhere Sony has a bunch of patents around RayTracing as well. I reckon Microsoft will be paying a lot more to AMD per console than Sony, this will all remain to be seen.

Yeah, but you don't need to virtualize the OS to do that. It would make more overhead.
It could be that the kernel uses only 80MB, but you have to take into account the caches and other structures used by it, which eat much more ram than that.
the PS5 "OS" or the user interface sits inside a virtualized container, this can be suspended and resumed on the fly every time you request
The OS can't be suspended, the user interface could be. But you don't need to run it inside a container to do that.

UNIX has a monolithic kernel, but Windows does as well (Windows NT is monolithic with some benefits of microkernel). So Microsoft could do the same as you are suggesting. I doubt neither would be doing it anyway.
You're kind of changing the roles of OS and games here. The OS is the one that must have the control to decide what runs and what not.
An OS that allows a program to take over the system has failed in its basic function of managing the system resources.
If you let developers control which application/process can or cannot be used when their game is played, expect the worse.
The OS moves programs or processes out of memory when they aren't needed, it has been like this since the beginning, but an address space reserved for OS is necessary, so the kernel uses it without fighting with other programs for memory. Letting the kernel and user space fight each other for memory leads to a big penalty in performance.
 

hyperbertha

Member
Nov 24, 2018
2,225
7,283
655
The ps4 can handle extremely detailed assets as seen in many games. You aren't looking at all of them close up at once. Same applies for next gen. The problem is that it seems current games are only using a fraction of currently available memory which is likely part of the reason you still see blurry textures and many low priority objects at significantly less detail. For more detail see further down this comment.

I'll quote myself
Yes I've always suspected current gen games are drastically limited by memory limitations in lowest common denominator hardware.
My question is what exactly is it that affects framerate other than resolution though? I mean won't trying to render a super high poly object cause a hit to performance other than just memory? If a GPU is constantly rendering frames that has extremely high poly objects even if they do fit within memory limits they should cause a decrease in framerate if my current understanding is correct.
 

Three

Member
Oct 26, 2014
6,708
4,526
690
I suspect that this will not change game design in a lot of the lower budget games. They will aim to appeal to as big an audience as possible so will not tailor their game to SSD drives but to last gen consoles and standard HDD on PC.

It will be interesting to see how the media and public reacts to this this time. At the start of last gen we had numerous outlets lambast games for 'not having a next gen experience', 'what's next gen about it' and slowing down next gen adoption. We had outlets saying 'you can't see differences anyway' . Lets see how it plays out this time with games that can offer next gen experiences but don't.
 

RCU005

Member
Nov 9, 2016
714
1,348
590
I would say in a properly designed game it will even make things worse.

This sentence sums up the reason why Mark Cerny said that they are testing PS4 games on PS5. This is why he said that PS5 might be too fast to run PS4 games. Sadly, people don't get it, and just jump into conclusions without any knowledge, evidence or anything.
 
  • Like
Reactions: psorcerer
D

Deleted member 775630

Unconfirmed Member
Certain people are downplaying the impact of SSD is there a reason? I thought Cerny talk about that was clear yet some people are still confused.
It's more about downplaying the impact of 5.5GB/s compared to 2.4GB/s SSD. The impact won't be twice as big.
 

psorcerer

Banned
May 1, 2012
2,479
3,537
1,035
Yes I've always suspected current gen games are drastically limited by memory limitations in lowest common denominator hardware.
My question is what exactly is it that affects framerate other than resolution though? I mean won't trying to render a super high poly object cause a hit to performance other than just memory? If a GPU is constantly rendering frames that has extremely high poly objects even if they do fit within memory limits they should cause a decrease in framerate if my current understanding is correct.

Yes and no. It depends on how you design your workload.
I can give you a simple example: tech demos. Why tech demos look that good? Because all the RAM and GPU power is used only to render that demo and nothing else.
Fast data streaming aims to achieve the same perf balance for every game. I.e. you should have in each frame only the data that's used for that frame. Nothing else.
 

RCU005

Member
Nov 9, 2016
714
1,348
590
It may end up producing a completely different approach to game design, or it may become a complete dud, its all dependant on how difficult it is to optimise for and if its worth the time, which will largely depend on how many units Sony sell, but also Microsoft has tried to go down a similar route, its a lot more standard pc based, but its inclusion does give 3rd parties an onus to at least attempt to change game design...

I think one thing we must consider, is that Sony first party business is very important for them. When designing a console, they tend to think more about what innovative ways THEY could do. It's similar to how Nintendo design their consoles and controllers: they think of themselves first, third party second. Microsoft on the other hand, are just starting to build their own first party business, so when they design a console, it's easier for them to make one that most developers will have a more standard approach.

If Microsoft begins to show full potential in first party games and becomes priority to them, we could start to see new console design based on innovations that work for them in the first place.
 
  • Like
Reactions: vdopey

Shmunter

Member
Aug 25, 2018
10,764
24,390
835
Yes and no. It depends on how you design your workload.
I can give you a simple example: tech demos. Why tech demos look that good? Because all the RAM and GPU power is used only to render that demo and nothing else.
Fast data streaming aims to achieve the same perf balance for every game. I.e. you should have in each frame only the data that's used for that frame. Nothing else.
Every gameplay scene looking like a cutscene is my outlook

Walking sim, cinematic platformer, open world. All of them.
 
Last edited:

Goliathy

Member
Mar 18, 2020
440
1,425
340
I think one thing we must consider, is that Sony first party business is very important for them. When designing a console, they tend to think more about what innovative ways THEY could do. It's similar to how Nintendo design their consoles and controllers: they think of themselves first, third party second.
What are you talking about? Where exactly is Sony doing this? Any example of a game that is not on PC/Xbox possible but running on PS4 right now?

Or are you talking about this:


Dat snow deformation though
 
Last edited:

Inviusx

Member
Jan 4, 2016
2,205
2,636
540
Reading this thread has made me appreciate how much technical wizardry went into making Doom Eternal so fucking smooth on console. Plus fast loading times, detailed textures, high resolution.

It's crazy.