Can the M60 be used for general purpose CUDA usage?

Accelerated Computing CUDA CUDA Programming and Performance
1

I see that the M40’s target audience is machine learning and that for the M60 the target is VDI.

Is there any reason why the M60 would not be good for general purpose CUDA usage?

I can see that it has a slower base clock speed (899 MHz as opposed to 948 MHz on the M40) but really that doesn’t seem like that big of a deal.