The problem with big data is that there’s a lot of it.
This is a trivial point, but it’s like asking what’s hard about lifting 495lbs. The 495lbs are the problem.
Every twelve second data run I take is 84mb. Each model degree of freedom increases the data size by a factor of a hundred, as in I can search for one mdof with 100 data runs, but two takes 10,000, and three mdof takes a million.