Thanks for heads up on D@H.
As to the FX cards I think it has to do with their Compute Capability of 1.1 and their on-board memory. Take a look at wikipedia's CUDA and Quadro pages along with nVidia's CUDA GPUS page. Those FX cards are of the same GPU series as the 8400M which is based on the Geforce 8xxx. Note that CUDA apps on the GeForce series requires 256MB of on-board memory but most mobile graphics use shared system memory which may be the problem.