PS2 performance analyzer statistics (from Sony)

[function]

Everything is exactly how it should be.

Obviously from a much more amateur point of view, that's kind of what I was thinking.

 

[Grall]

Thanks Faf, Deano and other PS2 devs for providing GOOD infos on the topic!

You all deserve some serious respect for sharing with the rest of us mere mortals.


*G*

 

[cthellis42]

The 15 - 20 million pps performance was being banded around here as an average figure representative of the performance of modern games.

Offhand, I don't think anyone has had bandied those numbers about as average numbers for ANY of the platforms. Just the occasional "performance snippit" that's all we can try to ween from the occasional developer comment. Only ever highlights "possibilities" though, and never a "performance aggregate." After all, the screen won't ALWAYS be filled with max enemies and explosions galore--or that would be one helluva game!

 

[No-name]

Yeah, that 15-20 million wasn't exactly a number that got peaked too much often in general, only few games here and there reached such numbers. I mean if games kept getting such numbers than it wouldn't be such an achievement now would it?

 

[marconelly!]

The 15 - 20 million pps performance was being banded around here as an average figure representative of the performance of modern games.

I don't think anyone was saying that, I know I didn't.

I would like to hear some peak numbers from the Performance Analyzer, if the numbers quoted in the first post are indeed average what certain games perform. 7.5M average is much different than 7.5M peak.

 

[function]

Hmm, this has got me thinking about which figures are for peak, average or sustained polygon counts.

The slide marked "rendering analysis" (13) states 10,000 and 145,000 polys per frame as being the minimum and maximum (or 600K to 8.7 million /s). Would this be 145,000 peak, average (over a section of the game) or sustained (over successive frames)?

As a later slide lists 125K per frame as being the "fastest so far @ 60fps", does this mean 125K [7.5 mpps] is the highest sustained or peak figure at 60 fps, but there are different results [e.g. higher polygon counts per second] for 30fps (where physics/AI/game state overheads may be less)? Or does it mean that 125K is the highest sustained, but 145K is the peak for any single frame generated for any game running on the PA?

If so, this would mean that no game running on the PA has ever hit 10 million pps (which is not to say the PS2 can't "do" 20 million pps).

Much of this is meaningless in terms of comparing consoles directly, but I'm just interested to know specifically what the figures are referring to. For the sake of knowing.

[Edited for clarity]

 

[nAo]

I would like to hear some peak numbers from the Performance Analyzer, if the numbers quoted in the first post are indeed average what certain games perform. 7.5M average is much different than 7.5M peak.

In our engine with very small not clipped triangles on the most used surface (2 vertex palettized colors per vertex, no normals/lighting, single textured) we reach about 22-25 MTriangles per second on batch of some hundreds triangles.
Average figures abruptly change with the amount of the on screen stuff, from 2 MTriangles/s up to 8-9 MTriangles/s. The highest average figures are reached on scenes with lot of small triangles rendering as much as 140000 triangles per frame (PAL, 50 fps)

ciao,
Marco

 

[cthellis42]

Much of this is meaningless in terms of comparing consoles directly, but I'm just interested to know specifically what the figures are referring to. For the sake of knowing.

Exactly, there's just too damn little out there to sate our "inner nerd."

 

[passerby]

Thanks to the people who replied. Question again. How successful do you feel you will be at utilizing VU0 in future and upcoming titles? Or will it be neccessary for the application's art/design/requirements to be "friendly" to VU0 usage for us to see effective usage of that component? Foe e.g. if a title's art design is such that it needs little particle effects, transparencies, etc. then VU0 will probably sit idle.

 

[archie4oz]

Well it's usage (or lack there of) in a studio's particular title may not be dependent on art assets. A lot can depend on a particular game's implimentation...

 

[Crazyace]

Measuring different things

One thing to note about the PA scans is that they often show entire game frames - poly numbers sometimes look less flattering due to multiple renderers, and idle time.
A game dev will often benchmark isolated engine elements, ( as shown above ) , giving the higher figures.
Sony demos ( even some ps2linux demos on the public site ) show high numbers - but a lot of games use more complex tcl code , or draw bigger polys, maybe even 'waste fillrate' on multiple full screen effects.

Lot's of factors affect graphics quality on consoles - sometimes just fitting all geometry and textures in memory in the first place is the biggest problem.

 

[cthellis42]

I imagine cache missing is still a bid thing no matter what, though.

 

[llemarie]

There is one thing I was wondering about - the app with 40x overdraw, is it a big secret which is it?

I'm not sure I'm supposed to say it, but I see no harm in it, especially that the game is doing pretty well technically. It was LOTR2, it is full of giant particles.

do you mind if I ask where you pull your information from? Individual hand-testing of every game through the PA yourselves only, or is there a lot of developer info-swapping going on as well?

Over the last two year we have scanned a large number of games. There is a developer community too, but they have their secrets that I cannot discuss it here.

Do you test only the maximum performance a title can bring into play, or average it with all the various levels a game can deliver? (Thinking specifically of racing titles here, as you can play with just yourself on a track, or potentially load it up with many more cars--meaning much more needed pushing power.)

I usually take a random point in the game. It might be unfair, but it is unfair in the same way for each game, so on average the results are still relevant.

I must say your "maximum" count is going to get a lot of people jumping around, but it's supposed to read more like "maximum average polys at 60fps for the entire title," right?

Reading the document without attending the talk is not so good, because of course you don't get the explanation for each assertion.
For example, the maximum count is indeed an average and is nowhere near the maximum the PS2 can achieve at peak.
If people like big numbers, a developer has contributed some code running a 600K poly textured scene at 60fps. It is public too, and can be found on playstation2-linux.com

I rather doubt the majority of ANY games for any of the consoles push over 5M at this point

I think so too. In most cases though, it is fair to say that the game has been designed not to need more that that, rather than it being limited by the hardware.

BTW: I loved your comments on the Data Packing slide talking about palletized textures

Thanks. During the talk I asked if there were any artists, I didn't expect any, but there were several of them actually. Busted.

the PA actually underestimates the polys drawn

Wrong, the PA shows you _exactly_ the number of polygons drawn, regardless of how many you actually intended to draw.

How does this affect your averages?

If affects the stats indeed. But in the end if you send 100K polys but only 50K got drawn, then you did only draw 50K. I think it is fair to count only polygons that are drawn on screen.

You are working for the developer support at SCEE (like listed in you presentation) this results in the following questions:

I am working for the developer support at SCEE, this results in the following answer:
No comments, sorry.

Hope this helps,
Lionel

 

[llemarie]

Missed a few questions, sorry:

Would this be 145,000 peak, average (over a section of the game) or sustained (over successive frames)?

145k sustained it was, but not at 60fps. That makes it 70K polys at 60 on average, so 125K at 60 is still the fastest (and those are actual displayed polygons). But those are just numbers, in most games the quality of the picture does not depend so much on the number of polygons.

I'm just interested to know specifically what the figures are referring to

I read the figures by looking at the rendering part of the scans only. That is to say I didn't include the waiting time for the VSync for instance. I did include the time waiting for texture uploads on the other hand, that could make a renderer capable of 15MPs run at 10 or 8 or less.

How successful do you feel you will be at utilizing VU0 in future and upcoming titles?

As we said, using VU0 is not easy. Results will certainly vary, we are still waiting to be surprised.

As a comment, it has been mentioned that when I said "VU0 usage" that wasn't taking VU0 as COP2 in account. That is true, and during the talk I did explain that the figures were for VU0 running independantly from the CPU, which was the relevant point I wanted to make.

 

[McFly]

Thx a lot for all the awnsers.

So it's LOTR2. Just curious, how did ZOE:TSR compare to LOTR2 (overdraw)?

I was very impressed with the animated textures in Silent Hill 3. Did you test that part of the game or just a random part of the game and if yes what was the conclusion?

Fredi

 

[llemarie]

Sorry Fredi, I am not allowed to discuss the results of the performance analysis on public forums. I don't even discuss them with developers, apart with their own games.

 

[cthellis42]

Thanks for answering and I hope you stick around... I'm sure you'll be getting a lot more. I'll just toss out a quick one:

If people like big numbers, a developer has contributed some code running a 600K poly textured scene at 60fps. It is public too, and can be found on playstation2-linux.com

Is this code that can be applied well in games, or more "demo code" that exists on its own right to show off capabilities but would be too hard to harnass in conjunction with all the other equirements a game would have to run as well?

 

[Nick Laslett]

With games like GTA:VC & SSX3 using VU0 to decode 4.0 DTS in real time, how does this appear in the statistics? I assumed that this would use a lot of the capabilities of VU0, or else they would decode 5.1.

It would appear that there is still a long way to go before the PS2 is fully utilised, if I am understanding the results of the PA sessions correctly?

Do the large number of technically average games released for the PS2 depress the results or have you only selected for analysis top tier games?

How does procedural rendering figure in this analysis? We heard a lot about this technique early in the PS2's life time, do many games use this technique and does it offer tangible benefits?

I believe games like Dropship, WRC and SSX use PR, do you know of any other notable examples?

 

[McFly]

Sorry Fredi, I am not allowed to discuss the results of the performance analysis on public forums. I don't even discuss them with developers, apart with their own games.

Damn.

Fredi

PS: No problem at all, was just interested.

 

[Legion]

With games like GTA:VC & SSX3 using VU0 to decode 4.0 DTS in real time, how does this appear in the statistics? I assumed that this would use a lot of the capabilities of VU0, or else they would decode 5.1

Is there any way the PS2 could encode DTS 5.1 much in the way the xbox encodes DD and output the digital signal through the SPDIF port?