Need For Speed Most Wanted Directx 9.0c Download
Hilke Mcnally
I've played NFS 5 a lot in my childhood on GeForce FX5700. Competed it fully 2 or 3 times. I've got only small issues: well known cpu speed related bug (it wasn't too annoing for me). Some fonts were weird but still readable. Ah, and also game crashed often in one place of one race of all game (i don't know why, but i replayed it some amount of times until successfully run it through).
My build back then was:
- Asus motherboard with socket 478 and Intel i865g chipset
- Intel Celeron D 2.8 GHz
- Asus GeForce FX5700 128 Mb
- 2 x 512 Gb DDR 400 RAM @ 333 MHz (because of limitation of CPU bus)
- Windows XP SP2
- DirectX 9.0c
- I don't remember GeForce driver version but it was the latest for mid-2000's
I got a Matrox Millennium II (MGA-2164W) today and wanted to test its capabilities. Since this is an early 3D accelerator, my expectations were fairly low. However, the card did surprise me in one aspect, as you will see below. All tests were performed under Windows 98SE with DirectX 9.0c.
> Another thing to be aware of is those Intel GMA cards have very
> obsolete specs and won't support anything better than Shader Model
> 2.0, raising Access violation exceptions on a bunch of DirectX 9.0c
> targetted games, like Need for Speed Most Wanted and newer, Command
> and Conquer Tiberium Wars runs with some tweaks and many other games
> don't even start. I recommend you use the most oldschool stuff you
> have if Intel's graphics crap is your target hardware :P
There are two speed tricks with Gaussian filtering using the pixel shader. The first is that the Gaussian filter (along with the box filter) is separable: you can filter horizontally, then vertically (or vice versa, of course). So for a 99 filter kernel you then have 18 texture samples in 2 passes instead of 81 samples in a single pass. The second trick is that each of the samples you use can actually be in-between two texels, e.g. if you need to sample texels 1 through 9, you could sample just once in between 1 and 2 and use the GPU to linearly interpolate between the two, between 3 and 4, etc., for a total of 5 samples. So instead of 18 samples you could get by with 10 samples. This is old news, dating back to at least ShaderX2 and GPU Gems, and we talk about it in our 3rd edition around page 469 on.