| yes, there are some experimental opcodes for running in the GPU, which are written in CUDA. There is a makefile there that
will be build the opcodes on OSX if the cuda toolset is present (for Linux it should be simple to modify it).
Performance in my MacBook GPU is OK, but in general not faster than the CPU. This is expected as I am sharing it with the graphics
etc.. I wanted to try a dedicated GPU to see what kind of performance I can get out of it.
Victor
On 15 Jan 2014, at 12:40, Matti Koskinen wrote:
> hi,
>
> I installed csound 6.02 on ubuntu 13.04, and as I was in need for a better RNG for windows, the rand() on mingw and VS is really not enough for another project, searching the source was the first place to look. I saw that there's a directory named cuda, and I'm wondering, are there really a cuda-gpu boosted things in csound, or are they in a development phase? Only my win7 machine has an Nvidia-card, Mac Mini has too, but only 256Meg stolen from main memory, and Opencv-cuda on win7 needs to be compiled with VS :-(
>
> Also got a new Android-tablet, Lenovo something ,7" one, which my wife bought for her old mother, but she can't see clearly a small display of it, so I confiscated it and installed csound.apk. Some csds I tried, look like working ok. Have to check the docs, what can and what can't be done with the Android-port. But seems to be interesting.
>
> tnx
>
> -m
>
>
>
> Send bugs reports to the Sourceforge bug trackers
> csound6:
> https://sourceforge.net/p/csound/tickets/
> csound5:
> https://sourceforge.net/p/csound/bugs/
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>
|