I am very sorry that Poem@home close its GPU project.
At the same time I understand from their website news that they are going to make available their source code.
Would be interesting if TNGrid would investigate if its projects could benefit from the implementation of such code

regards
Gabriele

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

"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)."