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

Status
Not open for further replies.
Yeah. IIRC "Primitive Shader" was mentioned by Mark Cerny when he was talking about the Geometry Engine in the Road to PS5 vid. Primitive Shader was introduced by AMD if I am not mistaking, during the Polaris or Vega Architectural reveal. Can't remember which exactly.
I think it's pretty safe to assume that the Geometry Engine in the PS5 does not have the hardware changes AMD made to allow support for Mesh Shaders. This may be the thing Microsoft is talking about when they say they waited for the full RDNA 2 feature set.
 
First, "Geometry Engine" is AMD's name for the fixed function geometry hardware in their GPUs. This isn't a Sony term.

Second, an older version of the Geometry Engine could simply mean support for AMD's Primitive Shaders/Primitive Discard without support for Mesh Shaders.

Third, I remember reading an article about Unreal Engine 5 and Nanite and one of the Unreal Engine devs specifically mentioned that any geometry not being rendered through Nanite would utilize the primitive shader path on the PS5. He did not say "mesh shaders" but specifically mentioned "primitive shaders" and specifically mentioned "PS5".
yea the quote is
"The vast majority of triangles are software rasterised using hyper-optimised compute shaders specifically designed for the advantages we can exploit," explains Brian Karis. "As a result, we've been able to leave hardware rasterisers in the dust at this specific task. Software rasterisation is a core component of Nanite that allows it to achieve what it does. We can't beat hardware rasterisers in all cases though so we'll use hardware when we've determined it's the faster path. On PlayStation 5 we use primitive shaders for that path which is considerably faster than using the old pipeline we had before with vertex shaders."

imo, if you are going to talk about a pipeline change -- that's fairly equivalent to what mesh shaders do as well; at the very least not the old front end.
 
Last edited:

alright.. maybe a comparison is happening here. between WD1 and WD4
edit: there was a patch. So perhaps he's just checking out any differences since the patch. Perhaps Discord is not recording his games correctly.

summary:
4k30 + raytraced reflections
higher quality render assets (textures, cube maps, geometry)
improved shadow quality
VRR
reduced loading times
oh, oh, did he say "stable 30fps on series x" in that video ... what does that mean for current gen consoles, if next gen consoles get promoted with stable 30fps?
 
It would be really interesting if Navi 21 and Navi 22 end up having some minor differences that reflected XSX and PS5. Although could such differences ended up hidden by the drivers?
There was a pretty transparent difference in the GPUs that launched around the time of the consoles, if Cerny's claim that the wider set of ACEs in Hawaii was lifted from the PS4. Actual feature set differences for API-related functions may be papered over by drivers, but would AMD want to add that level of disruption?
Some differences, like varying machine-learning ISA extensions for sub-families have been generally relegated to the specific markets that cared about them.

There were some Sony patents that described dividing screen space up into areas to be rendered at different resolutions to accelerate FOVeated rendering. I recall this being done at the geometry processor level rather than the pixel shader level like VRS. I could see a more VR focused Sony perhaps wanting to spend time customising an offshoot of Geometry Engines to support this, rather than waiting to take a newer geometry processor whose changes they didn't value as highly.
That specific part of the pipeline would seem like a subset of Tier 2 VRS, omitting Tier one and several other options in Tier 2.

Mesh shaders seem like a fantastic concept to me, but then again I have no idea what Sony's capabilities are, and what they might have given up but gained instead.
Sony cited primitive shaders, which AMD has backtracked on since failing to introduce them with Vega and paring their claims significantly for RDNA1. DX12 has gone more in Nvidia's direction, and while primitive and mesh shaders are generally situated in similar parts of the pipeline, their features and emphasis vary.
What Cerny mentioned did seem like it included features that AMD said might be possible with primitive shaders at their introduction with Vega--or some future version of primitive shaders different from mesh shaders.


I'm aware that this is what Anandtech has been told, but it's not what's been happening in practice. These are the roadmaps shown during 2019:

Both Zen 3 and RDNA2 are releasing their first models before 2021.
As such, it's reasonable to expect RDNA3 on 5nm to release before 2022, even if AMD don't want to compromise on that.
In practice, AMD's been burned repeatedly by promising tighter time ranges and having to delay. This era of meeting promises came after a change in policy to be more conservative so that AMD would underpromise and overdeliver, as opposed to the opposite.
Zen 3's release date is November 5 2020, or one respin away from being pushed into 2021.

I'm confused about this talk about PS5 "Geometry Engine" as if it's unique in itself. Is this not a generic term?
Sony does have a history of calling something $Generic_term Engine. AMD's used the term geometry engine for some of its GPUs in the GCN era, but there were multiple such blocks in GPUs like Polaris. A singular block in an RDNA-era GPU sounds like it might be in the same place as what AMD calls the Geometry Processor.
Cerny mentioned primitive shaders in the context of that new functionality, but not primitive shaders as AMD ever demonstrated. It's not clear if they fully correspond to what mainline RDNA would support.
 
Hypothetically speaking, if the performance of the two consoles is broadly similar on release how are we going to reconcile that with these comments from executives stating that the differences will be obvious and the PS5 isn't "full RDNA2"?
 
Hypothetically speaking, if the performance of the two consoles is broadly similar on release how are we going to reconcile that with these comments from executives stating that the differences will be obvious and the PS5 isn't "full RDNA2"?

Hypothetically, by simply believing that Sony money-hatted the developers for forced parity... :runaway:
 
Hypothetically speaking, if the performance of the two consoles is broadly similar on release how are we going to reconcile that with these comments from executives stating that the differences will be obvious and the PS5 isn't "full RDNA2"?
well.. it's likely to be a nothing burger I think.

But comparisons of games over time, and with PC and Xbox nearly fully aligned at the hip level for feature sets, makes the process easier for DF to identify gaps and spreads if they are present.
For instance - you build a game using DX12U feature set.
DF compares PC and XBox to see if they can find the equivalent settings
Using those settings, you should be able to figuratively figure out PS5 power and therefore resolution
If it's around the mark you guessed, it's a nothing burger.

If there is a wide discrepancy between expectation and actual, perhaps it's a RDNA 2 feature set thing.
 
Is that for real? If so, it is very interesting but I tend to be skeptical of unknown twitter users and their info.
The patent describes a method that divides screen into small sections and individually controls the render resolution of each section. This is much more coarse grained and it’s done at the expense of image quality, but it’s okay for VR because human eyes only see a very small area in the view sharply.

Also it works on different aspect than VRS: this controls raster rate(probably sampling rate), VRS controls shading rate. VRS respects polygon boundaries, this disregards that, which means if every triangle is exactly 1 pixel big VRS is useless (unless on deferred rendered parts), but again if triangles are that small hardware rasterizer will be next to useless too and this won’t help either.
 
Was hoping for more detailed info. Kind of a nothing burger :(
I was hoping for an "example we're the only one that supports VRS......."
Not just the general feature list.
Which I still have no reason to believe that some at least isn't on PS5. So if that's the case the list itself is meaningless for what I want to know.

In Princess Leah's voice
Digital Foundry your our only hope
 
Last edited:
So, the technique of increasing the resolution by ML is a reality on Xbox Series. Hm :)

Cages comments do need context. The interview reads more like specs theory crafting than something based on actual development. My headline from the interview:

"David Cage thinks that the PS5 will have better performance due to its stable development environment" :D
 
Does the PS5 not have any ML capabilities? If not, that's definitely a huge potential win for the Xbox hardware. Will be interesting seeing this generation pan out.
 
Does the PS5 not have any ML capabilities? If not, that's definitely a huge potential win for the Xbox hardware. Will be interesting seeing this generation pan out.
Kepler and GCN have ML capabilities :), CPUs regularly run ML as well.

Its mainly a question of software (in both quality and efficiency) and horsepower to run it; I assume what most people refer to as capability.

The latter is yes for the most part at least enough to support 30fps targets. I suspect 60fps will be tighter and unlikely (for both)
 
Last edited:
Status
Not open for further replies.
Back
Top