A note on down dating the cholesky factorization
The advantage of this approach is that there is a substantial saving of computer storage because the orthogonal matrix Q is not required.Convenient techniques are given to determine the search directions and Lagrange parameters that are needed in active set algorithms, and to revise the matrices of the calculation when the active set is changed.The system of claim 25 wherein said computation cells further include programmable latency delay means for delaying transfer of data between cells in order to achieve proper interleaving of said data, said configuration data further includes latency programming data, said programmable latency delay means responsive to said latency programming data to set said delay means in accordance with said latency programming data.
I was unable to find anything in the docs and also here in the issues, so I would really appreciate any information. PS: If this has not been implemented here, I would have to implement it for my use-case and if contributions on such a functionality would be welcome, I would be happy to make a contribution here (although I would be limited to python/cython).
For what it's worth, I completely agree that the ideal solution moving forward would be to have it integrated into LAPACK itself.