Re: vector & matrix operations
- From: "Robin" <sanchrp@xxxxxxxxxx>
- Date: Wed, 30 Mar 2005 19:49:34 +0200
All these are done on the CPU, GPU commands are carried out in Vertex
Shaders and Pixel Shaders. If the commands are done on a per vertex per
frame basis, they should be done in a vertex shader otherwise its better to
do them on the CPU, and pass the results as constants, saving the GPU from
doing the same calculation hundreds times. The same goes for pixel shaders,
though the time loss is much higher...( and all of this depends on whether
your targetted hardware supports shaders...)
Robin
"DPR" <temp@xxxxxxx> a écrit dans le message de news:
uxXpwmUNFHA.3844@xxxxxxxxxxxxxxxxxxxxxxx
> Hi,
>
> I just wondered if the vector and matrix operations (like D3DXVec3Add and
> all matrix transformations) are hardware accelerated or not. I couldn't
> find
> any information about this in the DirectX SDK. Which processing unit is
> responsible for these calculations ? The GPU or the CPU ?
>
> If it's the CPU - is there a possiblity to use the GPU instead ?
>
> Regards,
>
> --
>
> DPR
>
>
.
- References:
- vector & matrix operations
- From: DPR
- vector & matrix operations
- Prev by Date: Re: vector & matrix operations
- Next by Date: Re: vector & matrix operations
- Previous by thread: Re: vector & matrix operations
- Next by thread: Re: vector & matrix operations
- Index(es):
Relevant Pages
|
Loading