The quadro uses an diffrent architecture, the optizimation of this GPU is important.
For example for a short period of time the 780 has been faster in rendering than a 980, only because the 900 series was not yet optimized eventhough the 980 is much stronger. (by now 980 is much faster, opt. )
CUDA is distributed computing, quadro and tesla GPUs are all around CUDA. Quadro always have the most extended functions for CUDA, but in day to day use, these extra functions will not be called, but some people need them, and they will pay a lot for it.
For Blender Don’t go with Quadro, eventhough i asked the Q :), i did my research
Just go with a gaming card, since games do vector mult. and Matrix Mult. all the time, just like Blender does when rendering. So the workload is about the same just that blenders shaders are much much more complex creating a nicer render at time cost. (Comparing to an game engine)