Questions about PS2

Trying Googling about VU0. Sony said ages ago that something like only 10% of games used VU0 (or 10% utilisation, or somesuch) and PS2 had untapped potential (walled off through impossible complexity and obscurity as I understand it!). Didn't find the quote myself though.
 
Shifty Geezer said:
Trying Googling about VU0. Sony said ages ago that something like only 10% of games used VU0 (or 10% utilisation, or somesuch) and PS2 had untapped potential (walled off through impossible complexity and obscurity as I understand it!). Didn't find the quote myself though.
Click to expand...
Wasn't it for first, second, maybe third wave games. Maybe in last games or in at least some of last games siuation was different?
 
Liandry said:
So, polygons go from EE to EDRAM, then from EDRAM to GS and thei they are textured, and then for some effects they go to EDRAM again and from it to GS and they textured again. Right?
Click to expand...
EE to GS which renders polygons to EDRAM using textures which reside in EDRAM, as far as I know no polygon data is stored to EDRAM.
If you do multipass effects you send new polygon data trough EE to GS and re-render polygon into same location in EDRAM just with different source data. (also might write new texture to source alpha bit and use it when rendering next pass.)

If I remember correctly GS had 4MB buffer in which all read/write operations took place, meaning your textures, buffers, palettes etc didn't really have any difference it was just data which you could read/write freely.
Liandry said:
As I know it can only calculate geometry and lighting.
Click to expand...
Anything polygon related from generation to preparation for GS, yes. (Procedural geometry, surface evaluation/tesselation, instancing, animation etc.)
 
Shifty Geezer said:
From the days when we had good developer contributions to this forum. :(
Can't find anything about more recent usage.
Click to expand...
As I can see fom fist PDF from first link, there is 8% usage of VU0 and 56% usage of VU1. But it is first two years of PS2, I really think in he next years ysage for both VU's was a lot higher. I just look at games from first two years and games from 2005, 2006 on PS2. :D
 
I feel the whole EE-VU1-VU0 debacle is what jumpstarted modern day "secret juices" conspiracy theories.

It was a great time to be a gamer.
 
Shifty Geezer said:
https://forum.beyond3d.com/threads/ps2-performance-analyzer-statistics-from-sony.7901/

From the days when we had good developer contributions to this forum. :(
Can't find anything about more recent usage.

edit: Original presentation : http://www.gamepilgrimage.com/sites/default/files/SystemSpecs/PS2/HowFarHaveWeGot.pdf
Click to expand...

To be fair about that PDF, many, many games came out after it and the knowledge of how to use the PS2 had spread much further. I think it'd be illuminating if someone got ahold of those old tools and examine shit tons of games that came out after that PDF was published.
 
I.S.T. said:
To be fair about that PDF, many, many games came out after it and the knowledge of how to use the PS2 had spread much further. I think it'd be illuminating if someone got ahold of those old tools and examine shit tons of games that came out after that PDF was published.
Click to expand...
I also think the same. If there were some numbers for some games from 2005-2006, that woud be great.:D
 
London-boy said:
I feel the whole EE-VU1-VU0 debacle is what jumpstarted modern day "secret juices" conspiracy theories.
Click to expand...
Quite possibly. The first example I can consciously remember in the way of "hidden power"/"dark silicon" here on B3D is the Wii, where some people (Nintendo diehards, of course) were convinced there had to be untapped potential. It was explained by an unexpectedly large GPU as I recall.

Well, if there was anything hidden in that stupid little box it went to its grave with that power still hidden and locked away. Lol...
 
Liandry said:
Can anyone explain difference between Pixel Output and Pixel Fillrate. Because I thought it's the same thing.
Click to expand...

They're the same thing. Sometimes when you see the term "fillrate" by itself it refers to texture fillrate instead of pixel fillrate.
Liandry said:
As I can see fom fist PDF from first link, there is 8% usage of VU0 and 56% usage of VU1. But it is first two years of PS2, I really think in he next years ysage for both VU's was a lot higher. I just look at games from first two years and games from 2005, 2006 on PS2. :D
Click to expand...

Another source is cottonvibes, one of PCSX2's developers, who said in a 2011 blog post that VU0 usage is so low that emulating it with interpretation rarely has a significant speed hit:

The EE actually works very closely with VU0 as opposed to VU1, and so threading VU0 would not be a speedup because the EE would end up reading back from it too much. The good thing is that VU0 is rarely a bottleneck in games; this is evident because you can usually run VU0 interpreters and get a minimal speed-hit (if you try to run vu1 interpreters on the other hand, your speed will usually crawl to ~2fps).
Click to expand...

http://pcsx2.net/developer-blog/89-threading-vu1.html

This is far from an exhaustive statement but if VU0 usage was high in later popular games he probably wouldn't have made it.
 
Exophase said:
They're the same thing. Sometimes when you see the term "fillrate" by itself it refers to texture fillrate instead of pixel fillrate.
Click to expand...
Ok. Then we look at this presentation - http://www.gamepilgrimage.com/sites/default/files/SystemSpecs/PS2/HowFarHaveWeGot.pdf
At page 16 there is info about 3.6 Mpixels output. From 0.9 Mpixels to 12 Mpixels. With a maxium 40 Mpixels, for those times. But PS2 maximum is 1.2 Gpixel filrate. So how those number can be so low? Because even 40 (maximum for those times) is a lite bit more than 3%. :oops:

Exophase said:
This is far from an exhaustive statement but if VU0 usage was high in later popular games he probably wouldn't have made it.
Click to expand...
Then how was possible all those great games with great graphics, like GT4, GOW2, Black and many more others. I just can't belive what all this was done only on EE main core + VU1.
 
Liandry said:
Ok. Then we look at this presentation - http://www.gamepilgrimage.com/sites/default/files/SystemSpecs/PS2/HowFarHaveWeGot.pdf
At page 16 there is info about 3.6 Mpixels output. From 0.9 Mpixels to 12 Mpixels. With a maxium 40 Mpixels, for those times. But PS2 maximum is 1.2 Gpixel filrate. So how those number can be so low? Because even 40 (maximum for those times) is a lite bit more than 3%. :oops:
Click to expand...
They mean that 12Mpixels is 40 times the number of pixels in a then-typical output frame.

12Mpixels is referring to the most pixel fill during the rendering of a frame.

If a game does that while rendering at 30fps, that's 360 Mpixels, which is nearly 1/3rd of the theoretical peak fillrate.
 
HTupolev said:
They mean that 12Mpixels is 40 times the number of pixels in a then-typical output frame.

12Mpixels is referring to the most pixel fill during the rendering of a frame.
Click to expand...
Oh, I see now. I was a little bit confused here.

HTupolev said:
If a game does that while rendering at 30fps, that's 360 Mpixels, which is nearly 1/3rd of the theoretical peak fillrate.
Click to expand...
And here also. :D
But is those numbers for single pass? If there would be 2 passes, there woud be 720 Mpixels?
 
Liandry said:
But is those numbers for single pass? If there would be 2 passes, there woud be 720 Mpixels?
Click to expand...
No, that's total. But think about what's happening here. A screen has 0.3M pixels. Drawing 12 million pixels in one frame means drawing 40x as many pixels as there are on the display. That means the equivalent of 40 passes, not two. With the average being 12 times. So that's a typical 12 passes (assuming a pass is a full screensworth of pixels). From another perspective, 4K has 8 million pixels per screen. PS2 was typically drawing one and half 4K screens worth of pixels, breaking the final image down into lots of separate layers each drawn on top of the previous work.

The 1.2 GPixel fill is peak possible. We never achieve those values in actual software. That's what you could achieve if you didn't spend time calculating what those pixels should look like and just drew white pixels over and over. ;)
 
Shifty Geezer said:
No, that's total. But think about what's happening here. A screen has 0.3M pixels. Drawing 12 million pixels in one frame means drawing 40x as many pixels as there are on the display. That means the equivalent of 40 passes, not two. With the average being 12 times. So that's a typical 12 passes (assuming a pass is a full screensworth of pixels). From another perspective, 4K has 8 million pixels per screen. PS2 was typically drawing one and half 4K screens worth of pixels, breaking the final image down into lots of separate layers each drawn on top of the previous work.
Click to expand...
I though GS capable of only 4 pases. So it's really can do a lot more or not? I just shocked now.

Shifty Geezer said:
The 1.2 GPixel fill is peak possible. We never achieve those values in actual software. That's what you could achieve if you didn't spend time calculating what those pixels should look like and just drew white pixels over and over.
Click to expand...
I know that. :D Just used this number as an exampe.
 
You must log in or register to reply here.

Similar threads

S
SDL_Gpu API, SDL 3.0 ... New cross-platform graphics API.
Replies
1
Views
1K
Lurkmass
L
A
Understanding game performance through profiler trace graphs *spawn
2 3 4
Replies
66
Views
3K
XboxKING
X
D
Ray/Path tracing and 6th Generation games. Possible on Current Gen Consoles?
Replies
19
Views
3K
OlegSH
O
X
Shimmering lines, common in menus, HUDs: Term for the artifact? And possible to "fix" with post-processing?
2
Replies
26
Views
3K
Davros
D
D
The Complete List of PC Ray Traced Titles with Classification
16 17 18
Replies
359
Views
53K
DavidGraham
D
Back
Top