I am trying to prioritise the linking of nvblas over blas to work with Armadillo. From what I have read, I can see that the libraries need to be put in the following linking order: Lapack NVBlas Blas
I have done this in my visual studio project, however the matrix multiplication is not running on the GPU. Can anyone help me with this problem? I am using Visual studio 2019, Windows 10 and Armadillo. I have also tried armadillo with mlpack.
