Looking for multi-GPU applications

We are looking for OpenCL applications that were written with the intention of running them on several GPU devices in parallel and can run on Linux.

Any suggestions/pointers please?

Thank you very much.

Not a complete solution, but there is a multi-gpu tutorial for n-body simulation here:

http://www.browndeertechnology.com/docs … -rev3.html

There are simpler ways to play with metrics for multi-GPU scenarios, what do you need it for?

Thank you very much, we need this in order to evaluate the performance and network-overheads of running real-life applications on GPUs of remote computers, eg. on a cluster of computers with GPU(s), the more the better, because one can only place that many GPUs on a single computer.

More applications are most welcome.

Amnon, be aware that standard OpenCL only supports multiple devices within the same system. In other words, multiple GPUs or CPUs in the same computer. Network transparency is not a feature of OpenCL.

Dear David,

I am very aware of it, which is why we wrote an OpenCL library-platform that provides access to OpenCL devices (CPUs/GPUs) on as many computers across the network as you want. It works nicely with our synthetic applications, but we just lack the real-life applications to test it.

You may find useful this: http://www.luxrender.net/wiki/LuxMark

It has been successfully run on systems with 3xHD6990, 4x580GTX, etc.: http://www.luxrender.net/wiki/LuxMark_Results