A parallel implementation of the Karmarkar algorithm using a parallel Linear Algebra library
An important decision in the construction of a parallel Linear Algebra library is the choice of the process structure. The structure that is presented in this article allows for simple functional specifications of the processes and for their compositionality. In the functional specification of a process, matrices and vectors occur as ordinary variables and nothing is said about the communication behaviour of the instances of a process. The representations of the matrices and vectors are distributed across the process instances. A number of distributions are introduced. All library processes conform to the same representation details and communication conventions. The library processes can be composed sequentially, without requiring global synchronisation between process calls. As an example, a parallel implementation of the Karmarkar algorithm is briefly discussed.