Poem@home is going to make available their source openCl GPU code
would be interesting to understand if could be used also in A@H projects

The GPU accelerated algorithm will be open-sourced so that other groups can use the efficient forcefield freely to conduct research on GPU clusters on their own (the timeframe for this will be until Christmas. The code is ready to be shipped and the paper is written, but needs to be peer-reviewed before).

http://boinc.fzk.de/poem/