c++ matrix multiplication with LAPACK, BLAS, Boost uBLAS, or other efficient library

Project is to multiply large matrices as efficiently as possible in c++. For instance: A = 500,000 rows by 500 columns B = 500 rows by 700 columns C = 500,000 rows by 700 columns Compute: C = MatrixMultiply( A, B ); I need both the code and instructions to link the libraries involved. I work from Visual Studio 2013 on a Windows 8 machine with 16 GB memory. Graphics card is not CUDA, but I will upgrade soon. The project will be completed successfully when I have working code on my computer that reads a CSV from file (one for each matrix), then performs the efficient matrix multiplication. I have several years experience with c++ as a hobbyist. I'm hoping that with a little bit of hand holding, I can get this job completed very quickly. Thank you for your bid.