PGR3 (photo mode) pics

[standing ovation]

Let's face it.

As far as reality goes, being able to capture its essence -- let alone set it into motion -- will probably require a computing device that is billions upon billions of times more powerful than Cell ever though of being.

But IBM's new chip might be able to dodge those crippling calculations by manipulating motion pictures ...

 

[Lysander]

 

[overclocked]

I just laid out a my view on the renderer, garage is the besy looking thing in the game after the photo-mode(im not a cargame enthusiast but i would rather have the order of that split to when driving). As for the style and tracks/art i think its pretty bad, im more into RR/NFS if i had to choose, liked the latter one(DEMO) better on the PC with i find odd regarding the controller scheme.

 

[Nesh]

Uh...because one is still pictures, and one has to be...you know, actually calculated in real-time?

So?Models are the same, textures seem the same, and the console can display images at higher resolutions.

Its not like GT4 where the models, resolution and texture detail were lower for real time moments.

 

[Martin Eddy]

I think the motion blur puts all those high res textures to waste.

 

[Phil]

The best GFX in the game are found the the garage, this truly amazed me. the in-game stuff was a big step down for me. I look farward to the day when the in-game GFX will be this good:

And here we were arguing about those "developers-blog-screenshots" and all the excuses on why there were little to no screens that showed gameplay footage looking that good....

I must say I do like the lighting though.

 

[mckmas8808]

And here we were arguing about those "developers-blog-screenshots" and all the excuses on why there were little to no screens that showed gameplay footage looking that good....

I must say I do like the lighting though.

But some of those are from gameplay.

 

[mckmas8808]

More pics

More beautiful pictures of the best looking racing game in freaking history!

http://img.photobucket.com/albums/v217/antd/ruff2.jpg
http://img517.imageshack.us/img517/8241/191rq.jpg
http://mxz.lv/bildes/pgr1.JPG
http://mxz.lv/bildes/pgr3.JPG
http://mxz.lv/bildes/pgr4.JPG
http://img402.imageshack.us/img402/5105/141qt.jpg

 

[Bigus Dickus]

On a second note, who is beginning to think that the only serious difference between xbox360 and PS3 will be complexity of physics calculations, with the PS3 being cable of rendering a significant enough # of physics body dynamics/ simulation points over the xbox360 to make it noticeable to the end user (poor sentence structure eh? ).

I'll throw in my 2c and say that I'm not convinced that the increased theoretical power of the PS3 (vs. 360) will be able to be realized in perceptably better physics and/or AI. I've read a lot of discussions on the raw power of the respective processors involved, and it isn't obvious that the Cell architecture is as well suited to physics and AI as the XeCPU architecture. Perhaps it is, perhaps it isn't, but it is far from a given and we aren't talking about an order of magnitude difference anyway... so whether the processor differences are actualized as noticeable gameplay differences is far from certain.

 

[pakotlar]

I'll throw in my 2c and say that I'm not convinced that the increased theoretical power of the PS3 (vs. 360) will be able to be realized in perceptably better physics and/or AI. I've read a lot of discussions on the raw power of the respective processors involved, and it isn't obvious that the Cell architecture is as well suited to physics and AI as the XeCPU architecture. Perhaps it is, perhaps it isn't, but it is far from a given and we aren't talking about an order of magnitude difference anyway... so whether the processor differences are actualized as noticeable gameplay differences is far from certain.

considering how small the difference between some xbox1 and xbox360 games is with regards to the power gap, i would consider this a reasonable standpoint.

 

[Mintmaster]

I'll throw in my 2c and say that I'm not convinced that the increased theoretical power of the PS3 (vs. 360) will be able to be realized in perceptably better physics and/or AI. I've read a lot of discussions on the raw power of the respective processors involved, and it isn't obvious that the Cell architecture is as well suited to physics and AI as the XeCPU architecture. Perhaps it is, perhaps it isn't, but it is far from a given and we aren't talking about an order of magnitude difference anyway... so whether the processor differences are actualized as noticeable gameplay differences is far from certain.

A few of the rabid PS3 supporters here would disagree, but I totally agree. I've programmed a physics engine, and I can tell you that a 2-3x faster processing power (which Cell won't have in practice anyway) won't make a noticeable difference in gameplay, especially with some processing tasks varying with objects as N log N, N^2, and even N^3. Even more limiting is that FP power often isn't the limiting step, and latency-limited tree navigation is the problem. On a PC you'll notice it because the physics engine is tuned for one specific speed, so 1/3 the CPU speed makes for 1/3 the framerate. On a console, though, you tune your object number and physics accuracy to the processor, so gamers won't know what they're missing. Cross platform titles may show a small difference, but nothing 99% of gamers could notice.

In the end it will all come down to programming skills. I'd rather have better coders for my platform than 10x the CPU power. Graphics is a somewhat different story, but coders and especially artwork will still make an enormous difference.

Here's discussion I had regarding this issue:

I never said it wouldn't, but we're not talking about no physics vs. awesome physics. My point is 2 or 3 times the FP power (doubtful) will not have a significant impact on physics, especially in terms of what it brings you in a game.

In the mainstream physics realm, anything less than 10x more power isn't really noticable.

I've actually coded a physics engine. Yes, there's a lot of FP code in there, but really impressive physics (which I haven't achieved yet) needs to navigate and maintain large spatial data structures, solve sparse matrices, and know how to leave things alone that aren't going to change. It's not simply a matter of crunching numbers.

You're not kidding, there are some significant programming issues in physics, there are a lot of algorithims that would run 4-8x faster if hardware could support efficient scatter and gather of single FP(SP or DP) quantities into packed data structures. The problem is that the basic memory technologies are moving in the other direction.

 

[one]

I've programmed a physics engine, and I can tell you that a 2-3x faster processing power (which Cell won't have in practice anyway) won't make a noticeable difference in gameplay,

How do you define 'faster processing power'? For AGEIA's Physx chip, there are some cool demos which are not possible on Pentium, do you think it's not at all related to gameplay? Or cloth simulation or facial simulation?

 

[mckmas8808]

In the end it will all come down to programming skills. I'd rather have better coders for my platform than 10x the CPU power. Graphics is a somewhat different story, but coders and especially artwork will still make an enormous difference.

Before you read this I'm pretty sure you know Sony has good coders, but I have to say it. Sony may have more power (even though ease of coding levels that out) but Sony also has some of the best devs the world has to offer. So they probably aren't too worry about that.

 

[DemoCoder]

I'd personally take 10x the CPU performance *AND* good coders. There is no law that says that good coders and higher CPU performing platforms are can't simultaneously exist.

Just because an algorithm scales by N log N or N^2 rather than N, doesn't mean higher performance won't matter, it just means that algorithmic performance should be maximized first, then tweak to HW.

Amdahl's law doesn't say one should not care about performance.


On the non-local memory access front, as I pointed out in a past thread, this is not neccessarily a roadblock for CELL performance, it just means that physics code can't be as "general purpose" as one would like. The integration step is highly streamable, and collision detection contrary to common wisdom (some were claiming it is inherently serial which it is not) can be distributed.

The real burden is dealing with large databases. But some classes of algorithms, especially particle systems, should absolutely fly on CELL and certainly will be noticable. Actually, CELL + mega alphablending eDRAM GPU would be incredible for particle effects.

 

[Lysander]

http://www.msxbox-world.com/forums/index.php?showtopic=3890

 

[Martin Eddy]

http://www.msxbox-world.com/forums/index.php?showtopic=3890

Impressive.

 

[expletive]

In the end it will all come down to programming skills. I'd rather have better coders for my platform than 10x the CPU power. Graphics is a somewhat different story, but coders and especially artwork will still make an enormous difference.

Do better development tools factor into this at all?

 

[Lysander]

in-game pics from scaled at ten on 46" Samsung

 

[mckmas8808]

Nice! That's why I like that game.

 

[Laa-Yosh]

http://img262.imageshack.us/img262/1414/dscn32306sj.jpg

This could get away as a nighttime photo from a family trip to London...