Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[darius]

Well we already are using openGl so why not: This?

Basically it's a opengl shaders that are doing all the work. Of course that means that you need to have newer card but maybe 4 digit fps is (for gui) would be worth the trouble.
Now that i think about it i would rather see pathing implemented on shaders

[dorf]

Sounds awesome enough.
I wonder, if Baughn would be up for this. And, if it would be worth the extra hassle.

[Timst]

How ressource-heavy is the display system of DF anyway ? I always figured that the pathfinding was the main process.

[Lost Requiem]

Graphics power means soap to DF. The game is all about processing power, and shaders would be about as effective in DF as it would be in a Zork game.

[darius]

How ressource-heavy is the display system of DF anyway ? I always figured that the pathfinding was the main process.

Well 40D* series against 40D have a good increase in FPS. (e.g i was playing at about 70 fps with d16 it's 110)

Graphics power means soap to DF. The game is all about processing power, and shaders would be about as effective in DF as it would be in a Zork game.

You probably didn't look into it. It's about using shaders (or gpu for that mater) do computations and thus allowing cpu to do other things (like pathing).

[Baughn]

I don't think it would make much of a difference, but I might do it for fun. And then we'll see.

[Askot Bokbondeler]

I don't think it would make much of a difference, but I might do it for fun. And then we'll see.

fun? get back to work!

[dorf]

I don't think it would make much of a difference, but I might do it for fun. And then we'll see.

Thought as much.

[Freaky]

Now that i think about it i would rather see pathing implemented on shaders

This paper describes an efficient GPU implementation of parallel global pathfinding using the CUDA programming environment, and demonstrates GPU performance scale advantage in executing an inherently irregular and divergent algorithm.

[Draco18s]

Now that i think about it i would rather see pathing implemented on shaders

This paper describes an efficient GPU implementation of parallel global pathfinding using the CUDA programming environment, and demonstrates GPU performance scale advantage in executing an inherently irregular and divergent algorithm.

Except that not everyone has a GPU capable of taking advantage of CUDA.  You know, by being limited to a subset of all nVidia cards...

[Freaky]

Except that not everyone has a GPU capable of taking advantage of CUDA.  You know, by being limited to a subset of all nVidia cards...

So?  It's a research paper showing that pathfinding can work on a GPU, not a functional/usable implementation (rarely found in papers).  I have an ATI card, I'm well aware of the limitations of CUDA, it doesn't matter.  A proper library would probably use OpenCL, which can target multi-vendor GPUs and CPUs.

[Baughn]

6x speed boost in shader mode. 

[KillHour]

6x speed boost in shader mode. 

Holy Armok!  That's insane.

[Baughn]

Well, it just means there's no overhead from display anymore.

[KillHour]

Well, it just means there's no overhead from display anymore.

I know, but it's still impressive.  I assume that's VS OpenGL?