Matrix dimensionality problem in cvx

Currently I am working with an application, which involves optimization. Through simulation I observed that CVX is creating problem for large class of matrices (e.g matrix of dimension 1000 by 1000), it showing cvx_status as Inaccurate. So, I just want to know that is there any size limitation in CVX? And why it is showing cvx_status as inaccurate? Thanks in advance.

Have you tried another solver (for instance, both SeDuMi and SDPT3 under CVX)? Have you looked at the scaling of your variables and coefficients?

http://web.cvxr.com/cvx/doc/support.html#handling-numerical-issues