All geek questions in one place

OpenCl development on ATI and Nvidia at the same time

our working group is slowly trying a little OpenCl in a side project. So far, “everyone” is working on the NVIDIA Quadro FX 580. Now we plan to buy new computers for new colleagues, and instead of the FX 580 we could buy the ATI FirePro V4800 instead, which costs only 15 euros and gives us 1Gig instead of 512Gig Ram. which will be useful for our data intensity tasks.

So how difficult is it to develop OpenCl code simultaneously on Nvidia and ATI?

I read the next SO question, Running OpenCL on hardware from mixed vendors , which was very pessimistic about developing on / for different vendors. On the other hand, the question is already a year old.

What do you recommend?

+3
source share
4 answers

I used to work with the CUDA programming language.

I plan to start developing applications using OpenCL. As you mentioned, one of the best features of OpenCL runs on many hardware vendors (Intel, AMD, and Nvidia).

One project I came across that made extensive use of openCL for large-scale development, http://sourceforge.net/projects/hypgad/ . Perhaps it would be nice to take a look at the source code of this group and understand how they developed their application on so many hardware, including the sony cell processor.

PyOPENCL, , OpenCL, .

+2

? , , , .

, , , - CUDA/OpenCL , , ram, GPU , shoudd NVidia/AMD

0

- . ATI 64, Apple OSX 10.6, .

0

ATI NVIDIA , - SDK . OpenCL, OpenCL. (www.khronos.org/opencl), . - . , . , , - . (float2, float4) ATI, 4 ( RGB aplha).

0
source

Source: https://habr.com/ru/post/1772694/

More articles:


All Articles