I'm developing a Game with CS (as quickly as my daytime job permits, so please don't try to register for beta-testing yet ;-) and made this observation:
~100 FPS on my Athlon 2400+ Laptop with Mobility Radeon 9600, Single Channel 333Mhz RAM, CS0.98
~80 FPS on my Athlon 64 3500+ Barebone with Geforce 6800, Dual Channel 400Mhz RAM, CS0.99
Now what is with that? Is it 0.98 vs 0.99 ? Or Nvidia vs. ATI? Or is the 32bit Athlon more CS-friendly?
Basically I'm displaying 1 genmesh with ~130k polys and a single 1024x1024 32bit texture. so both graphics cards *should* do more than 100FPS anyway. Background is I try to do terrain, but for several reasons I do not want to use the CS terrain engine. So I create the terrain from a heightmap using the simplest, fastest thing there is: genmesh.
(If I'm wrong there please correct me!).
Any ideas or pointers will be appreciated.
Regards
Christian
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
If you want to know whether it's CS or your hardware, I suggest you test CS 0.99 on your Laptop as well as 0.98 on your barebone.
One guess: at some point we disabled vertex buffer objects(=a way to load vertex data directly into the system ram) by default in CS0.99 due some problems that it caused. Check the "Video.OpenGL.UseExtension.GL_ARB_vertex_buffer_object" setting in r3dopengl.cfg and try setting it to 'true'. You also try the inverse for 0.98 (setting it to false).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hmm, unfortunately I was forced to use 0.99 on the Barebone because 0.98 would crash.
Well, walktest and my app and half a dozen others would crash, awstest for example would run.
I'll look into 0.99 on the laptop and see if this makes a difference.
Just tried Video.OpenGL.UseExtension.GL_ARB_vertex_buffer_object = true, you're perfectly right about it causing issues-> crashed at once
The barebone hardware as such has been HL2 tested, it IS faster than the laptop ;-).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi There,
I'm developing a Game with CS (as quickly as my daytime job permits, so please don't try to register for beta-testing yet ;-) and made this observation:
~100 FPS on my Athlon 2400+ Laptop with Mobility Radeon 9600, Single Channel 333Mhz RAM, CS0.98
~80 FPS on my Athlon 64 3500+ Barebone with Geforce 6800, Dual Channel 400Mhz RAM, CS0.99
Now what is with that? Is it 0.98 vs 0.99 ? Or Nvidia vs. ATI? Or is the 32bit Athlon more CS-friendly?
Basically I'm displaying 1 genmesh with ~130k polys and a single 1024x1024 32bit texture. so both graphics cards *should* do more than 100FPS anyway. Background is I try to do terrain, but for several reasons I do not want to use the CS terrain engine. So I create the terrain from a heightmap using the simplest, fastest thing there is: genmesh.
(If I'm wrong there please correct me!).
Any ideas or pointers will be appreciated.
Regards
Christian
If you want to know whether it's CS or your hardware, I suggest you test CS 0.99 on your Laptop as well as 0.98 on your barebone.
One guess: at some point we disabled vertex buffer objects(=a way to load vertex data directly into the system ram) by default in CS0.99 due some problems that it caused. Check the "Video.OpenGL.UseExtension.GL_ARB_vertex_buffer_object" setting in r3dopengl.cfg and try setting it to 'true'. You also try the inverse for 0.98 (setting it to false).
Hmm, unfortunately I was forced to use 0.99 on the Barebone because 0.98 would crash.
Well, walktest and my app and half a dozen others would crash, awstest for example would run.
I'll look into 0.99 on the laptop and see if this makes a difference.
Just tried Video.OpenGL.UseExtension.GL_ARB_vertex_buffer_object = true, you're perfectly right about it causing issues-> crashed at once
The barebone hardware as such has been HL2 tested, it IS faster than the laptop ;-).