Stefano Salvini

Learn More
Developers of parallel applications can be faced with the problem of combining the two dominant models for parallel processing—distributed-memory and shared-memory parallelism—within one source code. In this article we discuss why it is useful to combine these two programming methodologies, both of which are supported on most high-performance computers, and(More)
This paper discusses methods for expressing and tuning the performance of parallel programs, by using two programming models in the same program: distributed and shared memory. Case studies show how mixing these two approaches results in efficient machine use because the two models match the two levels of parallelism present in the architecture of current(More)
The HiPerDNO project aims to develop new applications to enhance the operational capabilities of Distribution Network Operators (DNO). Their delivery requires an advanced computational strategy. This paper describes a High Performance Computing (HPC) platform developed for these applications whilst also being flexible enough to accommodate new ones emerging(More)
Alternating Direction Implicit (ADI) methods provide a computationally efficient way to solve for antenna based gains in full polarization. In this paper, we analyze the convergence of such methods in simulations. We show that convergence of a basic implementation can be quite slow and we propose two forms of relaxation to improve convergence behavior. The(More)
The HiPerDNO project aims to develop new applications to enhance the operational capabilities of Distribution Network Operators (DNO). Their delivery requires an advanced computational strategy. This paper describes a High Performance Computing (HPC) platform developed for these applications whilst also being flexible enough to accommodate new ones emerging(More)