• 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.

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

Status
Not open for further replies.

Bo_Hazem

Banned
Ye
Yep, I had the chance to test some high grade headphones... And they were something sublime (you need a proper amp and dac too). But considering that ppl like how beats sounds... I like how everyone is praising (rightfully? We'll see) sony audio solution and then same person will use TV speakers or beats cans...

Never touched those extremely professional headphones myself, but the difference from high quality MP3 and CD-quality is already massive. Sony using some magic to push the sound quality higher with their high-resolution tech that even CD's sound like shit in comparison:




Aiming for Sony's WH-1000XM4 if it's 100% compatible with PS5 🙌 Astro A40+mixer is just not packing enough bass punch so far but good quality overall so far.

 
D

Deleted member 775630

Unconfirmed Member
I always wondered how does this work with consoles that only have one hdmi port already taken by the TV
Two options (I think)
1. Console -> receiver -> soundsystem & TV
2. Console -> TV -> Soundsystem

In case of step 2, you normally also have HDMI outputs on your TV. HDMI ARC, audio return channel or something like that
 

draliko

Member
Two options (I think)
1. Console -> receiver -> soundsystem & TV
2. Console -> TV -> Soundsystem

In case of step 2, you normally also have HDMI outputs on your TV. HDMI ARC, audio return channel or something like that
yep, step 2 is preferred to avoid latency and sync problems (usually only the good amps got a real pass through channel on hdmi) but you will be limited to toslink specs (arc is a toslink replacement), with hdmi 2.1 (i believe) eARC was introduced, and it's the best choice if avaiable.

wwjgKZ0.jpg
 
Last edited:
Wow.. This was unexpected. Looks like Sony is pushing it even further by having some custom transistors inside their GE.

It's called mesh shaders and both systems will support that. VRS and Mesh shaders are two different things that save performance in different areas, not sure why they are even being mentioned together. Both will be used together, it's not a one or the other thing.
 
Last edited:

RaySoft

Member
It's called mesh shaders and both systems will support that. VRS and Mesh shaders are two different things that save performance in different areas, not sure why they are even being mentioned together. Both will be used together, it's not a one or the other thing.
I took for granted both systems had the same functions regarding the GE. But then he's tweet don't make any sense? It's reading like he's comparing it to another solution.
Sounded like Sony's solution was that the primitives aren't even used if it's not needed. I guess this is where the cache scrubbers comes in.
If you know wich data is not needed already (cache scrubbers job) then you already know that you don't need to calculate the vertices either.. I don't think XSX have anything comparable to the cache scrubbers?
 
Last edited:
Where did you get this info from

The guy who made the tweets is this guy:

Rich Geldreich is a founder of Binomial, the company behind the Basis Universal Supercompressed GPU Texture Codec. And apparently Microsoft is using this technology in XSX.

https://github.com/BinomialLLC/basis_universal/blob/master/README.md

And apparently he is still working with MS to improve the compression format to match the 6GB/s (minus overhead ) compression of the XSX HW block.

Also I have read that the XSX's implementation of Direct Storage seems to use a special game format dedicated to SSD so both PC and XSX and can take advantage of the storage and compression benefits exposed in the DS API.

This may be one of the reasons why they require XSX enhanced titles to operate strictly from their specified SSD offerings and not USB externally connected devices.
 

Bo_Hazem

Banned
I guess MS just chose the term "virtual memory" because they thought it would be the easy way for most gamers to comprehend the concept? It's not a particularly accurate description though.

Nope, they've CHOSEN that term because they're lying and misleading, like with their FUD campaign that they have a long history with. Same happened with Project Acoustic vs The Tempest 3D Audio engine laughable comparison, they try their best to mislead people, but people shame Sony for being so technically transparent and deep. Like they're lying, MS, now that they don't care about the "power" narrative when it's been their main focus all along after Ali, Crytek dev, burst their bubble.

d79489ef86e5dd4524714b54fe86ac36.jpg
 

Gamernyc78

Banned
Ppl wanted clarification on psv bc and yup almost all the ps4 games or majority will be bc not what others were saying about it only being hundreds because Cerny mentioned certain bc games getting the boost treatment. Thousands so tht puts it as gd or better than the competitors bc 😊

 
Last edited:
I took for granted both systems had the same functions regarding the GE. But then he's tweet don't make any sense? It's reading like he's comparing it to another solution.
Sounded like Sony's solution was that the primitives aren't even used if it's not needed. I guess this is where the cache scrubbers comes in.
If you know wich data is not needed already (cache scrubbers job) then you already know that you don't need to calculate the vertices either.. I don't think XSX have anything comparable to the cache scrubbers?

I don't follow what you are saying there. Why would Sony's primitive shaders be cache dependent?
 
Last edited:

RaySoft

Member
Nope, they've CHOSEN that term because they're lying and misleading, like with their FUD campaign that they have a long history with. Same happened with Project Acoustic vs The Tempest 3D Audio engine laughable comparison, they try their best to mislead people, but people shame Sony for being so technically transparent and deep. Like they're lying, MS, now that they don't care about the "power" narrative when it's been their main focus all along after Ali, Crytek dev, burst their bubble.

d79489ef86e5dd4524714b54fe86ac36.jpg
I def. know where you're comming from, but it's not like they're straight up lying. I guess it's culture thing. US has a tendency of sugar-coat things so it's more pleasing. You can see it many MS statements for XSX as well.. i.e. the backwards compatebility.. MS straight out says "full" compat. while Sony were talking about they going through all the games to assure compatebility. It's the same thing, but to the common american that reads quite different. MS is just more forwardleaning (US style) while Sony are more open with their current status. Neither is FALSE for what they deliver in the end, but it's interperated quite different by people across the world.
 

RaySoft

Member
I don't follow what you are saying there. Why would Sony's primitive shaders be cache dependent?
That the cache scrubbers comes before the primitive shaders in the pipeline.. Why feed the primitives and waste cycles if it's not really needed. He's post hinted to that the XSX don't have theses scrubbers, so the engine don't know if it's needed or not when the shaders are fed their lunch. If you could stop the calculations at an earlier state, you wouldn't waste cycles feeding them their job (that is not even needed)
 
Last edited:

Bo_Hazem

Banned
I def. know where you're comming from, but it's not like they're straight up lying. I guess it's culture thing. US has a tendency of sugar-coat things so it's more pleasing. You can see it many MS statements for XSX as well.. i.e. the backwards compatebility.. MS straight out says "full" compat. while Sony were talking about they going through all the games to assure compatebility. It's the same thing, but to the common american that reads quite different. MS is just more forwardleaning (US style) while Sony are more open with their current status. Neither is FALSE for what they deliver in the end, but it's interperated quite different by people across the world.

I think it's a crime in the US to false claim things to sell products, but probably we're overwhelmed with false ads that they rarely get called out and sent to court. Not to mention that a huge corporate have an army of lawyers, if not twisted ways and bribes to protect them.
 
Then why in pc gpus do we not see higher clocked gpus perform better then lower clocked ones.

Exactly, there are limits to what higher clocks can do for you past a certain point. Also, if you look at some of the newer, more advanced techniques like Mesh Shaders or Primitive Shaders, what is it that they do?

They are meant to be far, far superior and more performant ways to handle geometry processing as well as culling. Why are they more suited for geometry compared to the old way of doing things? It's because mesh shaders and primitive shaders are more similar to the far more flexible compute shader model and can access things like group shared memory among other things.

It's primarily because of this that for the purpose of handling geometry it's now possible to better take advantage of the highly parallel nature of GPUs compared to the more limited, less performant way that it was handled before. The strength of a GPU is that they can spawn and spread work across many more cores more easily than a CPU can.

You see where I'm headed right? You look at the description on both nvidia's blog and the directx blog for something like Mesh Shaders, and it becomes clear pretty fast that for these newer, more advanced hardware features it's actually better to have a bigger, wider GPU with more compute units and shader cores, not a smaller GPU with less cores and simply a higher clock speed.

Mesh Shaders, Amplification Shaders for helping with the culling, even Ray Tracing, they all benefit more from having a bigger, wider GPU with more CUs.


The new shaders bring the compute programming model to the graphics pipeline as threads are used cooperatively to generate compact meshes (meshlets) directly on the chip for consumption by the rasterizerT. Applications and games dealing with high-geometric complexity benefit from the flexibility of the two-stage approach, which allows efficient culling, level-of-detail techniques as well as procedural generation.

The mesh shader stage produces triangles for the rasterizer, but uses a cooperative thread model internally instead of using a single-thread program model, similar to compute shaders. Ahead of the mesh shader in the pipeline is the task shader. The task shader operates similarly to the control stage of tessellation, in that it is able to dynamically generate work. However, like the mesh shader, it uses a cooperative thread model and instead of having to take a patch as input and tessellation decisions as output, its input and output are user defined.

he new mesh shader pipeline provides a number of benefits for developers:

  • Higher scalability through shader units by reducing fixed-function impact in primitive processing. The generic purpose use of modern GPUs helps a greater variety of applications to add more cores and improve shader’s generic memory and arithmetic performance.

That was from Nvidia's blog. Now check the DirectX blog


By bringing the full power of generalized GPU compute to the geometry pipeline, mesh shaders allow developers to build more detailed and dynamic worlds than ever before.

Prior to mesh shader, the GPU geometry pipeline hid the parallel nature of GPU hardware execution behind a simplified programming abstraction which only gave developers access to seemingly linear shader functions. For instance, the developer writes a vertex shader function that is called once for each vertex in a model, implying serial execution. However, behind the scenes, the hardware packs adjacent vertices to fill a SIMD wave, then executes 32 or 64 vertex shader functions in parallel on a single shader core. This model has worked extremely well for many years, but it is leaving performance and flexibility on the table by hiding the details of what the hardware is really doing from developers.

Mesh shaders change this by making geometry processing behave more like compute shaders. Rather than a single function that shades one vertex or one primitive, mesh shaders operate across an entire compute thread group, with access to group shared memory and advanced compute features such as cross-lane wave intrinsics that provide even more fine grained control over actual hardware execution. All these threads work together to shade a small indexed triangle list, called a ‘meshlet’. Typically there will be a phase of the mesh shader where each thread is working on a separate vertex, then another phase where each thread works on a separate primitive – but this model is completely flexible allowing data to be shared across threads, new vertices or primitives created as needed, existing primitives clipped or culled, etc.

Less CUs with a higher clock would be more beneficial to the PS5 if the intent was to use the old techniques. But if you're using these newer techniques, you actually want a bigger GPU. And keep in mind the Series X GPU isn't exactly slow in the clock speed department either. It's clocked faster than a RTX 2080 Super.
 

RaySoft

Member
I think it's a crime in the US to false claim things to sell products, but probably we're overwhelmed with false ads that they rarely get called out and sent to court. Not to mention that a huge corporate have an army of lawyers, if not twisted ways and bribes to protect them.
I have not done any fact checking but I'm pretty sure US has the most lawyers per capita.
 

Bo_Hazem

Banned
I have not done any fact checking but I'm pretty sure US has the most lawyers per capita.

It's a crime in the US, on paper at least:

"In the United States, there are state and federal false advertising laws that prohibit various types of deceptive advertising, misleading labeling, and similar practices. False advertising laws provide important rights for consumers, arming them with the ability to seek monetary damages when they've been misled."

 

RaySoft

Member
It's a crime in the US, on paper at least:

"In the United States, there are state and federal false advertising laws that prohibit various types of deceptive advertising, misleading labeling, and similar practices. False advertising laws provide important rights for consumers, arming them with the ability to seek monetary damages when they've been misled."

I think most countries in the world has the same stance on this, at least in europe. False advertising is quite bad. But ayways..
MS was just talking about the end product, and Sony were commenting on current status. Both systems will ship with more or less the same amount of compatibility, I guess.
 

Grodiak

Member
"In the United States, there are state and federal false advertising laws that prohibit various types of deceptive advertising, misleading labeling, and similar practices. False advertising laws provide important rights for consumers, arming them with the ability to seek monetary damages when they've been misled."

Semi-unrelated, but it is weird how bashing the competition in ads is, however, somehow ok in the states. That wouldn't fly in a lot of European countries. The first time I saw Samsung's "Ingenious" ads with apple products visible and even their logo - I almost spat my milk and cookies.
 
Last edited:

Sosokrates

Report me if I continue to console war
Exactly, there are limits to what higher clocks can do for you past a certain point. Also, if you look at some of the newer, more advanced techniques like Mesh Shaders or Primitive Shaders, what is it that they do?

They are meant to be far, far superior and more performant ways to handle geometry processing as well as culling. Why are they more suited for geometry compared to the old way of doing things? It's because mesh shaders and primitive shaders are more similar to the far more flexible compute shader model and can access things like group shared memory among other things.

It's primarily because of this that for the purpose of handling geometry it's now possible to better take advantage of the highly parallel nature of GPUs compared to the more limited, less performant way that it was handled before. The strength of a GPU is that they can spawn and spread work across many more cores more easily than a CPU can.

You see where I'm headed right? You look at the description on both nvidia's blog and the directx blog for something like Mesh Shaders, and it becomes clear pretty fast that for these newer, more advanced hardware features it's actually better to have a bigger, wider GPU with more compute units and shader cores, not a smaller GPU with less cores and simply a higher clock speed.

Mesh Shaders, Amplification Shaders for helping with the culling, even Ray Tracing, they all benefit more from having a bigger, wider GPU with more CUs.






That was from Nvidia's blog. Now check the DirectX blog




Less CUs with a higher clock would be more beneficial to the PS5 if the intent was to use the old techniques. But if you're using these newer techniques, you actually want a bigger GPU. And keep in mind the Series X GPU isn't exactly slow in the clock speed department either. It's clocked faster than a RTX 2080 Super.

Interesting stuff, its going to be really exciting see these new technologies come together.
 

RaySoft

Member
Semi-unrelated, but it is weird how bashing the competition in ads is, however, somehow ok in the states. That wouldn't fly in a lot of European countries. The first time I saw Samsung's "Ingenious" ads with apple products visible and even their logo - I almost spat my milk and cookies.
I think it's cool if it's done with some finesse.
 
Last edited:

Bo_Hazem

Banned
I'm hoping there is no 100GB partition for streaming, that sounds like a drive killer (similar to an SSD cache on a SQL server). Creating a lot of unnecessary writes, copying data from one part of the drive to another. I'm hoping the games are packaged so that the install itself is accessible by the CPU/GPU directly, this would eliminate wasted writes.

That's what duplicates mean, XSX needs 100GB worth of duplicates. PS5 does not need duplicates, streamed directly, because it's designed so.

PS5 works like this:

ps5-ssd-gdc-presentaion-2.jpg


XSX works like this, like a PC or slightly better:

148294.jpg


The highlights of PS5 advantages in the SSD field:

148295.jpg
 

RaySoft

Member
That's what duplicates mean, XSX needs 100GB worth of duplicates. PS5 does not need duplicates, streamed directly, because it's designed so.

PS5 works like this:

ps5-ssd-gdc-presentaion-2.jpg


XSX works like this, like a PC or slightly better:

148294.jpg


The highlights of PS5 advantages in the SSD field:

148295.jpg
Ehh... The XSX don't need any dupicate data either since that's the benefit of static nand chips contra platter drives with seek times. The only reason why they duplicated data on platter drives was to reduce seek times on platter drives. If the HDD head was at a different location that the data needed was, you had to wait for the head to move before you could reed the sectors. This is eliminated with SSD wich both consoles have.
 
Last edited:

Bo_Hazem

Banned
Ehh... The XSX don't need any dupicate data either since that's the benefit of static nand chips contra platter drives with seek times. The only reason why they duplicated data on platter drives was to reduce seek times on platter drives. If the HDD head was at a different location that the data needed was, you had to wait for the head to move before you could reed the sectors. This is eliminated with SSD wich both consoles have.

It's the 100GB partition that works like ancient gaming duplicates going forward. If that's not the case, games that are 100GB on XSX might end up being 20GB on PS5, so to avoid that they made this afterthought trick to compensate for its inferior SSD and I/O.
 
Last edited:

Jtibh

Banned
My 5 year old daughter wanted me to download the psone version of spyro after beating the ps4 version twice. Well with a little bit of my help.

Had to dust off my ps3 and dick around with updates and log ins etc.
I forgot everything about it and was very surprised of the little icons for each game being animated and each of them played a short track of the game on loop.

I want this feature back on next gen
 

RaySoft

Member
It's the 100GB partition that works like ancient gaming duplicates going forward. If that's not the case, games that are 100GB on XSX might end up being 20GB on PS5, so to avoid that they made this afterthought trick to compensate for for its inferior SSD and I/O.
Okay? I thiught this was part of MS' direct X suite? So any gamecode can just ask for a file, and the API gives it to you independant on if it's from a SSD or HDD? i.e delivery is transparant to the gamecode. (with exception of maybe giving the gamecode the info if the data comes from HDD or SSD, so the game can cope thereafter?
 
Last edited:

Space_nut

Member
Exactly, there are limits to what higher clocks can do for you past a certain point. Also, if you look at some of the newer, more advanced techniques like Mesh Shaders or Primitive Shaders, what is it that they do?

They are meant to be far, far superior and more performant ways to handle geometry processing as well as culling. Why are they more suited for geometry compared to the old way of doing things? It's because mesh shaders and primitive shaders are more similar to the far more flexible compute shader model and can access things like group shared memory among other things.

It's primarily because of this that for the purpose of handling geometry it's now possible to better take advantage of the highly parallel nature of GPUs compared to the more limited, less performant way that it was handled before. The strength of a GPU is that they can spawn and spread work across many more cores more easily than a CPU can.

You see where I'm headed right? You look at the description on both nvidia's blog and the directx blog for something like Mesh Shaders, and it becomes clear pretty fast that for these newer, more advanced hardware features it's actually better to have a bigger, wider GPU with more compute units and shader cores, not a smaller GPU with less cores and simply a higher clock speed.

Mesh Shaders, Amplification Shaders for helping with the culling, even Ray Tracing, they all benefit more from having a bigger, wider GPU with more CUs.






That was from Nvidia's blog. Now check the DirectX blog




Less CUs with a higher clock would be more beneficial to the PS5 if the intent was to use the old techniques. But if you're using these newer techniques, you actually want a bigger GPU. And keep in mind the Series X GPU isn't exactly slow in the clock speed department either. It's clocked faster than a RTX 2080 Super.

:)
 

CrysisFreak

Banned
My 5 year old daughter wanted me to download the psone version of spyro after beating the ps4 version twice. Well with a little bit of my help.

Had to dust off my ps3 and dick around with updates and log ins etc.
I forgot everything about it and was very surprised of the little icons for each game being animated and each of them played a short track of the game on loop.

I want this feature back on next gen
Legit Sony needs to bring that futuristic style back, everything needs to be damn smooth and interactive.
PS3's (and PSP's) XMB had kino elegance tbh.
 

CrysisFreak

Banned
I still had tlou on the harddrive and when the main theme kicked in just by scrolling on it made we want to start playing it.
Sometimes i need that extra motivation to get myself to play
Yeah a small animation window for the current game would take barely any system resources at all and it would definitely be cool.
They did announce there's a lot of OS features of PS5 that are new and unrevealed so let's see what they cooked up.
I just hope that it is not too iterative, need something more exciting and next-gen.
 

Jtibh

Banned
Yeah a small animation window for the current game would take barely any system resources at all and it would definitely be cool.
They did announce there's a lot of OS features of PS5 that are new and unrevealed so let's see what they cooked up.
I just hope that it is not too iterative, need something more exciting and next-gen.
Would love to have something like " previously on insert title" for when you drop a game for weeks and then you want to get back to it but cant remember what happened .
Would be cool
 
That's what duplicates mean, XSX needs 100GB worth of duplicates. PS5 does not need duplicates, streamed directly, because it's designed so.

PS5 works like this:

ps5-ssd-gdc-presentaion-2.jpg


XSX works like this, like a PC or slightly better:

148294.jpg


The highlights of PS5 advantages in the SSD field:

148295.jpg

I'm all for sony in this console war, but thats not how any of this works.

Both consoles will use ssds. Both consoles won't need duplicate data. So as long as the game developer cares they will be able to push the game without duplicate data to the consoles.
Both consoles won't have seek times for files and so on because they don't use hdds. ( not taking into account harddrives connected via usb )
 
All your posts have been Microsoft this and Microsoft that. You asked for information on what Sony is doing that is similar to what Microsoft was doing and i provided it to you.

Let me give you a 101 on what virtual memory is.

You have a main storage that is the SSD and the RAM.

RAM is usually very small in size (16GB) and usually very fast.

Applications (games) are very large in size.

The operating system transfers any data that is needed by the CPU and GPU into the RAM and when it needs more data, it overwrites old data with new data.

This is where virtual ram comes in play. Instead of overwriting the old data, it moves the old data to the virtual ram on the SSD and changes the address space to point to the SSD then a new data takes the space in the RAM.

If the CPU or GPU needs the data that was moved to the SSD, it first checks in the RAM and if it is not there it checks in the SSD, the data is then transferred.

All you have done is just sequestered a partition of your main storage to be used as virtual ram. The OS knows that there is 100GB SSD storage that it can use to store inactive data. The speed at which the data is moved is bound by the bandwidth of the storage. It is not a new thing. PS4 does this as well, it has an extra 1GB slower memory that it moves background applications to when it is not active.
Sounds a lot like RamDisk tbh.
 
Less CUs with a higher clock would be more beneficial to the PS5 if the intent was to use the old techniques. But if you're using these newer techniques, you actually want a bigger GPU. And keep in mind the Series X GPU isn't exactly slow in the clock speed department either. It's clocked faster than a RTX 2080 Super.
The number of CUs and frequency has absolutely nothing to do with how a CU processes data. The XSX's and PS5's CUs will do the exact same thing. You are mixing up the scale at which things happen. The old way processed things at the vertex level, which was really an abstraction since that's not how the GPU processed them. The GPU packaged up many vertices and processed them at the same time within a CU.

The new way of processing data gives the programmer control over how that processing of many vertices works inside the CU. What it does not do, and this is very intentional, is control what happens across CUs. Each CU is independent, which is absolutely necessary if you want efficient parallel processing. The last thing you'd what to do is create dependencies from one CU to another so that they had to be executed in one particular way.

So the number of CUs is completely irrelevant, by design, to the programming logic. The logic should not care if two packages of data were processed in two separate CUs at the same time, one after another on the same CU, or one after another on separate CUs.

Note #1: There likely are cache optimization strategies that don't make the CU work completely independent, but they don't prefer a slow wider path over a fast but narrower one.

Note #2: I'm not sure, but I think amplification shaders can create work for other CUs. However even if they do, once they've created it, each CU would still operate independently.
 

PaintTinJr

Member
Exactly, there are limits to what higher clocks can do for you past a certain point. ....
I really want take the hit and read all the stuff you’ve linked about discrete PC hardware – despite my prejudice that it isn’t relevant to HSA systems - to be respectful of where you are coming with your viewpoint, so we can delve into this difference of on a technical level about which configuration will suit the newer path-traced rendering paradigm better. But before I read all that, can you give a pseudo game code rendering workload – preferably Path Traced – where you can prove your opening statement? And I mean prove without needing some PC benchmarks to infer your assumption is correct, and instead explain a pseudo coding example - from your own understanding of 3D game rendering that makes your argument for you.
 
It's the 100GB partition that works like ancient gaming duplicates going forward. If that's not the case, games that are 100GB on XSX might end up being 20GB on PS5, so to avoid that they made this afterthought trick to compensate for for its inferior SSD and I/O.

Ridiculous. 🤣

You know the SSG implementation dates back to 2016 ON THE MARKET. This is the first appearance of the tech on console.

Carry on though
 

SonGoku

Member
Ye
Yep, I had the chance to test some high grade headphones... And they were something sublime (you need a proper amp and dac too). But considering that ppl like how beats sounds... I like how everyone is praising (rightfully? We'll see) sony audio solution and then same person will use TV speakers or beats cans...
Two options (I think)
1. Console -> receiver -> soundsystem & TV
2. Console -> TV -> Soundsystem

In case of step 2, you normally also have HDMI outputs on your TV. HDMI ARC, audio return channel or something like that
Any suggestions for HQ/Lossless audio connection for headphones in consoles, is the quality of wireless headsets like WH-1000XM3 up to par
Do hdmi eArc amps exist or just usb?
 
Last edited:

B_Boss

Member
XSX will remain on top, PS5 cant go beyond its established computational limits, it will just get closer to peak (higher VALU utilization) compared to XSX

I suppose this console generation has me thinking about (for fun and genuine curiosity) which console will reach its peak performance most efficiently, within their own engineered parameters if that makes sense 🤔?
 
Last edited:
The number of CUs and frequency has absolutely nothing to do with how a CU processes data. The XSX's and PS5's CUs will do the exact same thing. You are mixing up the scale at which things happen. The old way processed things at the vertex level, which was really an abstraction since that's not how the GPU processed them. The GPU packaged up many vertices and processed them at the same time within a CU.

The new way of processing data gives the programmer control over how that processing of many vertices works inside the CU. What it does not do, and this is very intentional, is control what happens across CUs. Each CU is independent, which is absolutely necessary if you want efficient parallel processing. The last thing you'd what to do is create dependencies from one CU to another so that they had to be executed in one particular way.

So the number of CUs is completely irrelevant, by design, to the programming logic. The logic should not care if two packages of data were processed in two separate CUs at the same time, one after another on the same CU, or one after another on separate CUs.

Note #1: There likely are cache optimization strategies that don't make the CU work completely independent, but they don't prefer a slow wider path over a fast but narrower one.

Note #2: I'm not sure, but I think amplification shaders can create work for other CUs. However even if they do, once they've created it, each CU would still operate independently.
Remember this come from the guy who said to exhaustion the PS5 doesn't has hardware RT because reasons

I will ignore that guy before I he start with his conspirations
 

PaintTinJr

Member
I'm all for sony in this console war, but thats not how any of this works.

Both consoles will use ssds. Both consoles won't need duplicate data. So as long as the game developer cares they will be able to push the game without duplicate data to the consoles.
Both consoles won't have seek times for files and so on because they don't use hdds. ( not taking into account harddrives connected via usb )
I think he might be referring to the engine data that will go into the 10GBs(effectively VRAM) of XsX. Mark explained a very simple setup of streaming just FOV data a full quality in the time it took the player to rotate. The split access pattern on XsX - along with slower SSD and no IO complex - may result in some of the old design style loading of non-visible data being needed to multi-plat PS5 design lead games, which in turn would require a drop in the FOV asset quality because the effective size of 10GB would be reduced by the out of view assets.
 

SonGoku

Member
Less CUs with a higher clock would be more beneficial to the PS5 if the intent was to use the old techniques. But if you're using these newer techniques, you actually want a bigger GPU. And keep in mind the Series X GPU isn't exactly slow in the clock speed department either. It's clocked faster than a RTX 2080 Super.
The advantages of a higher clocked smaller card are intensified with RDNA2 features since the Geometry Engine that handles these new techniques will perform faster
It will also help get closer to peak compute performance
 

PaintTinJr

Member
I suppose this console generation has me thinking about (for fun and genuine curiosity) which console will reach its peak performance most efficiently, within their own engineered parameters if that makes sense 🤔?
With Cerny saying they've done away with loading times, and that we'd like the cooling solution, it had me thinking. Nonsense, I know, but what if with a return to no loading times, we also got a passively cooled console - other than a large slow quiet fan for reducing the ambient environment temp to a constant,
 
Status
Not open for further replies.
Top Bottom