Orbital Transformation in MD

[August Melcher]

I also run into similar issues when running a restricted calculation with 
everything else the same. In the paper on second generation Car-Parrinello 
MD, they mention "Since the ASPC integrator preserves the idempotency 
constraint only approximately, it sporadically has to be explicitly 
enforced, either by a Cholesky decomposition or by very few purification 
iterations " could this have something to do with the slowdown I am 
noticing?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140514/76c0e894/attachment.htm>