🔒 Locked

Display List Efficiency

Started by eldee May 28, 2002 at 11:56 AM 0 replies 500+ views

Original Post

May 28, 2002 at 11:56 AM

Are display lists ever *less* efficient than drawing polygons manually? im certain that they become less efficient at some point since they get loaded into memory.. my question then is actually when do they become less efficient? I'm porting an old 2D tile engine i did in DirectDraw over to OpenGL.. i hope to load a tileset and store each tile as an element of a Display List. Max size of a tileset is 1024x1024 with a tilesize of 32x32. that'd be 1024 tiles... it'd probably be much less than this, but for saftey's sake im going to stick with my 1024 figure. would it be more efficient to just write a Tile Blitting function that coordinated the offset of the texture coords? -eldee ;another space monkey; [ Forced Evolution Studios ]

[edited by - eldee on May 28, 2002 1:00:39 PM]

invective

May 28, 2002 at 12:14 PM

For 2D games you are most certainly fillrate limited and not geometry limited, so forget trying to optimize the geometry -- its not going to make your game any faster. 2048 triangles at 60 frames a second is only 122,880 triangles a second, which is less that 1% of the 20 million theoretical triangles a second that even a GeForce 2MX can do. The nVidia ribbon demo runs at 16 million dynamic triangles per second over the AGP bus on my Geforce 2MX, so i seriously doubt that any geometry optimizations you make will have any quantifiable effect. Optimize only portions of your code that are bottle necks -- otherwise you are just wasting your time.

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.