Hello: I am wondering is double precision supported in current 4.6.1 GPU version? Otherwise it would be very slow to use CPU version running free energy calculations.... thank you very much best Albert