Agni's Philosophy runs at 60FPS on a GTX 680, uses 1.8GB VRAM. Can next-gen run it?

orioto

Good Art™
I'm not aware of any developer who builds levels 10x bigger than what current consoles can actually handle. So that's definitely going to take more time and goes back to the problem I'm speaking of.

And having better toolsets doesn't matter if games aren't being made faster. Well, you could hypothetically create current gen games faster, but is that the standard you want on hardware more capable than that?

Why would level be 10 times bigger.... Not every game has to be gta for christ sake. Not every game has to be virtual life. That's this kind of thinking that kills budgets.

Assets are already high poly then downgraded anyway. And Square has been producing hours of cgi for its ff for ages now. Even in the ps1 era, i guess they had to modelize, texturize and render all those backgrounds...
 
Doesn't GTX680 draw 140W+ on average? Can't imagine that a decent i7 and other current specs are great for power budget either

I'm curious to see what's actually possible to fit into an acceptable console sized box.
 

Dice

Pokémon Parentage Conspiracy Theorist
I just want proper AA that works on all detail mapping/tessellation, shadows, reflections, under DoF, etc. It's so rare. Why add stuff that will just look like a pixelated mess? Such a waste of power.
 

JordanN

Banned
Why would level be 10 times bigger.... Not evefy game has to be gta for christ sake. Not every game has to be virtual life. That's this kind of thinking that kills budgets.
I think you need to tell that to publishers and more importantly, gamers. Not me. :p

Edit: The second part may not disagree with what I said but I'll continue to watch.
 

orioto

Good Art™
I think you are simplifying things. Why are games costing more and more to make if everything in place is already done? The game industry wouldn't have needed to leave the era of PS1/N64 budgets apparently.

Good question! To me it's more the excessive content in games. I mean i don't know how square is handling that but solething like FFVIII or IX probably cost a lot to. When you think about it, the level of details for backround is way more important in those games... AND you have more and bigger towns... Something is wrong. To me the thing that really need more hands is creating character assets now, and monsters. But the number of monster models in FFXIII is ridiculously small... So ?? Would this gen engine only would have cost too much to develop ?? I wonder
 

Setsuna

Member
I really doubt that this will look like next gen, in fact I'd wager it won't look like this except for some Heavy Rain type experiences (if that)

Pre-scripted sequences is not a game, and a 680 is well above the tech next gen can afford.

all games are pre scripted
 

Zaventem

Member
Mismanagement most likely.When i look at the scope of something like Dark Souls and Demon's Souls it's insanely detailed. The i dunno beastiary? if you want to call it that.. hardly has obsessive model color swaps unlike FFXIII.I wonder what they were doing all this time.
 

Dice

Pokémon Parentage Conspiracy Theorist
I do feel like bigger towns are important if we're going to keep with our current trends. Skyrim tried so hard to feel epic, but when the apparently huge towns are populated by all of 20-30 people, it doesn't feel like saving a nation, you know? They try to get into all this political shit with the elves and talk of war and civil war and everything, but it's just small groups of dickheads. The bandits out in the wild seem to outnumber soldiers and citizens 10:1, but generally just everything related to people seems so small. In older times you say well hey man, we just have a couple spites on a screen. It's obviously an interactive comic book or something like that, so imagine bigger. These days, there are few excuses, especially with older games like Daggerfall having been designed like they were. We should be rethinking stricture and resource management for the whole experience, not just screenshots.
 
D

Deleted member 20920

Unconfirmed Member
I do feel like bigger towns are important if we're going to keep with our current trends. Skyrim tried so hard to feel epic, but when the apparently huge towns are populated by all of 20-30 people, it doesn't feel like saving a nation, you know? They try to get into all this political shit with the elves and talk of war and civil war and everything, but it's just small groups of dickheads. The bandits out in the wild seem to outnumber soldiers and citizens 10:1, but generally just everything related to people seems so small. In older times you say well hey man, we just have a couple spites on a screen. It's obviously an interactive comic book or something like that, so imagine bigger. These days, there are few excuses, especially with older games like Daggerfall having been designed like they were. We should be rethinking stricture and resource management for the whole experience, not just screenshots.

I still think that the bigger towns and larger NPC pools must come with equal content. I do not need these extra work done just for the sake of immersion. There is also the issue of manpower and costs. Those are very valid excuses. The towns and people within will always be a representation of the actual size of the city and population (within the context) and will never be fully realised so why go for excessiveness at the cost of having more unique content?
 
I just want proper AA that works on all detail mapping/tessellation, shadows, reflections, under DoF, etc. It's so rare. Why add stuff that will just look like a pixelated mess? Such a waste of power.

Image quality should definitely be a priority, I've played so many PC games with advanced graphical features that look like shit in motion thanks to jaggies, flickering shadows, pop-in etc.
Meanwhile a game with much less advanced graphics looks better simply cause it has really good IQ, it looks clean, sharp but most of all it has a good AA solution that works.
 

Shikoro

Member
is the entire city tessellated like that nvidia demo? If so, holy shit!! I can't believe they rid of LoD altogether ^_^ I really hope this leads to massive performance boosts to other areas like more eye candy or physics. LoD = sucks
One would hope so. LoD swapping is unacceptable in this day and age. I will not be buying games which don't use tessellation for what it's intended...
 
R.I.P developers next gen.

If that level of detail becomes standard, I can see many studios crashing. And that's not even populating such towns with people.

They are getting the models from Visual Works CG which is something they do in every game. I guess that not every scenario will have that excat level of detail but right now is not more expensive if we go by what SE is saying, actually the contrary.
 

Go_Ly_Dow

Member
I have a question, but aren't assets/models/textures always produced at a higher quality anyway (due to the power of the PC's they're made on) and then downscaled to run on consoles?

If that's the case, then hypothetically next-gen development costs won't be a huge leap over current-gen, since optimization should be simpler due more power to work with.
 

zallaaa

Member
Well i'm seriously thinking that you are a troll at this point
your opinions≠facts and so streets of rage2 isn't much better looking than skyrim not even in billions of years
You are avoiding some posts because you obviously cannot answer them
you cannot make a game with physics gameplay on the xbox or genesis
you cannot make a game with kinect gameplay on the xbox or genesis
you can make a good game without horse power
you cannot make any games you want without power
the lack of power limits innovation on gameplay
This is my last response directed to you because obviously you are either a troll or are entirely closed into your idea and you will not accept that someone says otherwise.

I believe his point is that you could downgrade any game to a level so simple that you could play it on any kind of hardware (somebody did port the mechanics of Portal to a TI calculator for example), so your point could be reversed. What he seems to miss is that you wouldn't be playing the same game at that point.
 
I don't get why people want ultra high resolutions or high framerates out of every console game. That's the one case in which I'd say, if you care about these things so much get a PC.
 
I have a question, but aren't assets/models/textures always produced at a higher quality anyway (due to the power of the PC's they're made on) and then downscaled to run on consoles?

If that's the case, then hypothetically next-gen development costs won't be a huge leap over current-gen, since optimization should be simpler due more power to work with.

A large amount of assets that use normal mapping nowadays (almost everything) are already made at a higher quality.

drake_zbrush2.jpg
 

Krabardaf

Member
Every asset that uses normal mapping nowadays (almost everything) is already made at a higher quality.

Not at all, most normal maps aren't generated from sculpt. That would take far too much time, even if it obviously yields better results.
This is almost always true for characters though.
 
Not at all, most normal maps aren't generated from sculpt. That would take far too much time, even if it obviously yields better results.
This is almost always true for characters though.
I only know about Doom 3s enviroment textures which have been modeled beforehand and it looked that way from some of the normal maps I've seen.

Saying "every asset" is not true though, yes, edited the post.
 

Krabardaf

Member
I only know about Doom 3s enviroment textures which have been modeled beforehand and it looked that way from some of the normal maps I've seen.

Saying "every asset" is not true though, yes, edited the post.

yes it happens and i'm all for hand sculpted environment, but there are many ways of creating normals maps, some are generated from the base texture, other are simple patterns(hate that, pointless), they can also be genrated from custom heightmaps.
The very nature of normal maps make them most efficient when generated from actual geometry, but producing high def models of everything would take an immense amount of time. Few, if any studios, does this.
 

Krabardaf

Member
Because 10m polys per frame means it's pushing 600m polygons per second...
Well polygons/seconds doesn't really make sense in videogames, per frame is what matters.
10M can be reached via tessellation, tessellated geometry cost is null for CPU, it's a completely different thing and the overall performance impact is, while different, far less.
 
Because 10m polys per frame means it's pushing 600m polygons per second...
Xenos supposedly is rated for 500 million polygons, whilst that is obviously much greater than is realistically possible, there's no reason to think that a GPU that is more than 13x faster won't be able to at the very least reach 100 million above the Xenos' theoretical max.
 

Dice

Pokémon Parentage Conspiracy Theorist
Image quality should definitely be a priority, I've played so many PC games with advanced graphical features that look like shit in motion thanks to jaggies, flickering shadows, pop-in etc.
Meanwhile a game with much less advanced graphics looks better simply cause it has really good IQ, it looks clean, sharp but most of all it has a good AA solution that works.
I have turned off so many features/switched to DX9 mode on so many games just to get a clean and solid image. I also bought Witcher 2 at launch and still haven't played it because AA didn't work and my card isn't powerful enough for ubersampling. I don't want my memories of that game to be a jaggy, shimmering mess. It's one thing to go back and play PS1 games, and something else entirely to have these problems when you know what it was intended to look like and instead are getting problems that were reintroduced to PC gaming.
 

i-Lo

Member
New 4Gamer Video. Good next gen chara create sliders, wao.


As impressive as they are (and they are astounding indeed) these character models are shown off in vacuum, i.e. without a surrounding that demands more polygons and textures be rendered. What guarantees do we have of character models being either that detailed or retaining that fidelity in game, especially, after the transition Lightning went through from the E3 video to the final game?

A very long article about the engine (all in Japanese)

http://www.4gamer.net/games/032/G003263/20121201006/

Someone mentioned that certain scenes would be pulling about 10 million polys in real time + texture. IIRC RSX in PS3 is capable of around 250 million poly/sec (I don't know if they're textured or not) which amounts to around 8.33 million polygons per 30th of a second in a 30 fps game. So, a jump of 1.7 million polys per 30th of a sec equates a total of 300 million polys/sec. Doesn't sound like a lot and was wondering if it's indicative of what sort of GPU power are we talking about for next gen...

is the entire city tessellated like that nvidia demo? If so, holy shit!! I can't believe they rid of LoD altogether ^_^ I really hope this leads to massive performance boosts to other areas like more eye candy or physics. LoD = sucks

Oh yes this. I am tired of pop-in. Aside from lack of AA and AF (on consoles), Pop in thanks to different LoDs has been the biggest bane of visual immersion. Next gen GPU should have powerful tessellation units.
 

i-Lo

Member
Because 10m polys per frame means it's pushing 600m polygons per second...

So since the days of RSX, can't today's GPUs push well past the 600mil poly/sec while textured?

Given that GPU's capabilities are today measured only in TFlops, it's a shame we don't get to learn about their poly/sec figure (in general). I wonder what kind of poly stats would a Radeon Cape Verde/Pitcairn produce today as opposed to RSX's 250 mil/sec figure...
 
Because 10m polys per frame means it's pushing 600m polygons per second...
For reference and if i recall correctly, there were scenes in a modified version of Crysis when i was playing that had 5,000,000 tris per frame. Those counts could be similar or bigger in Crysis 2 because there are some objects being heavy tesselated. I remember just some concrete barriers of close where segmentend in 1000,s of polygons.
 
600M triangles is NOTHING to modern day GPUs.

We are in 2012. GPUs are ridiculously powerful. They make super data crunching computers out of these things people.

600M triangles seems like a lot right?

http://www.behardware.com/art/imprimer/857/

The 680 can push 3.5 BILLION (w/ triangles not facing the camera culled) & 1.6 BILLION w/ no culling of triangles

7970 can do 1.5 billion w/o culling

these GPUs are MONSTERS
 
So since the days of RSX, can't today's GPUs push well past the 600mil poly/sec while textured?

Given that GPU's capabilities are today measured only in TFlops, it's a shame we don't get to learn about their poly/sec figure (in general). I wonder what kind of poly stats would a Radeon Cape Verde/Pitcairn produce today as opposed to RSX's 250 mil/sec figure...

http://www.behardware.com/art/imprimer/857/

7870 can do...

1.7 billion triangles drawn per second (none culled)
almost 2 billion triangles drawn per second (triangles culled that arent facing camera)

if you're wondering why the 7870 is higher than the 7970, it's because the 7870 is a 1GHZ stock chip vs a 925mhz chip. Both have the same number of ROP's

it absolutely destroys the gpus in the PS360
 

i-Lo

Member
http://www.behardware.com/art/imprimer/857/

7870 can do...

1.7 billion triangles drawn per second (none culled)
almost 2 billion triangles drawn per second (triangles culled that arent facing camera)

if you're wondering why the 7870 is higher than the 7970, it's because the 7870 is a 1GHZ stock chip vs a 925mhz chip. Both have the same number of ROP's

it absolutely destroys the gpus in the PS360

Aye, thank you credible hulk.

In a console environment, we'd probably see down clocked version of a pitcairn (lower than even 925Hz; perhaps around 850Hz). Even then the count should be well above a 1.1 billion (with culling). That should be plentiful.

It makes me wonder how best will the devs utilize this additional power given just how incredibly good character models have begun to look (especially in the twilight years of current gen) with smart design choices (in general).
 

RoboPlato

I'd be in the dick
Aye, thank you credible hulk.

In a console environment, we'd probably see down clocked version of a pitcairn (lower than even 925Hz; perhaps around 850Hz). Even then the count should be well above a 1.1 billion (with culling). That should be plentiful.

It makes me wonder how best will the devs utilize this additional power given just how incredibly good character models have begun to look (especially in the twilight years of current gen) with smart design choices (in general).
This is what I was thinking too. Especially for devs that have been working with modern GPUs on PC for a while (DICE, CryTek) a haddware update would allow them to integrate it into the already more optimized console targets and start taking advantage of the hardware more quickly. A lot of devs have learned a lot about optimization this gen and I hope that carries over.
 
characters will look even better, but i think they look pretty good right now.

environments will see the biggest jump in fidelity IMO.
 

i-Lo

Member
characters will look even better, but i think they look pretty good right now.

environments will see the biggest jump in fidelity IMO.

I think we're reaching the point where additional polygons will be best utilized for great vistas (draw distance, no more 2D BS) and eliminating pop-in (oh please please please) because as any self respecting graphic enthusiast would acknowledge that it's all about the lighting for realism.
 
Can you explain?

Because the way it was explained to me, it seems like tesselation can be used on everything - level geometry, characters, static objects, the whole shebang.

Maybe i got this wrong my impression was just use tessellation because we have it?
It has it uses but please not like Crysis 2 cement blockades. That was Tessellation because we can do tessellation and write that of the checklist.
 
Top Bottom