Efficient Algorithms for the Multicore and Heterogeneous Architecture: A Shortcut to the Efficient Utilization of the Multicore and Heterogeneous Architecture. Cet article n’est pas disponible.Efficient Algorithms for the Multicore and Heterogeneous Architecture: A Shortcut to the Efficient Utilization of the Multicore and Heterogeneous Architecture

High performance computing (HPC) plays a fundamental role in tackling the most intractable problems across a wide range of disciplines and pushing the frontiers of science. It now becomes a third way to explore the unknown world besides theory and experiments. Manycore architecture and heterogeneous system is the main trend for future supercomputers. The programming methods, computing kernels, and parallel algorithms need to be thoroughly investigated on the basis of such parallel infrastructure. This book is driven by the real computational needs coming from scientific and industrial applications. For example in different fields of reactor physics such as neutronics or thermohydraulics, the eigenvalue problem and resolution of linear system are the key challenges that consume substantial computing resources. In this context, our objective is to design and improve the parallel computing techniques, including proposing efficient linear algebraic kernels and parallel numerical methods.

High performance computing (HPC) plays a fundamental role in tackling the most intractable problems across a wide range of disciplines and pushing the frontiers of science. It now becomes a third way to explore the unknown world besides theory and experiments. Manycore architecture and heterogeneous system is the main trend for future supercomputers. The programming methods, computing kernels, and parallel algorithms need to be thoroughly investigated on the basis of such parallel infrastructure. This book is driven by the real computational needs coming from scientific and industrial applications. For example in different fields of reactor physics such as neutronics or thermohydraulics, the eigenvalue problem and resolution of linear system are the key challenges that consume substantial computing resources. In this context, our objective is to design and improve the parallel computing techniques, including proposing efficient linear algebraic kernels and parallel numerical methods.