High Performance Computing in Science and Engineering

Modern scientific computing depends on more than faster processors: algorithms, software, parallel architectures, and the applications they support all have to work together. High Performance Computing in Science and Engineering brings that broad landscape into view through nine peer-reviewed papers from the 2019 HPCSE conference.

The proceedings move from the practical challenges of datacenter operation to mathematical methods and computational applications. For readers working with scientific or parallel computing, the value is in seeing different problems—and different approaches—collected in one research-focused volume.

Research across the HPC landscape

The contributions address models, algorithms, and software for using modern parallel and distributed systems, with examples drawn from science and engineering. The collection includes work on thermal conditions in a Tier-0 datacenter, resilience in PDE frameworks and linear solvers, fast matrix-vector multiplication, and large-scale boundary element algorithms.

From numerical methods to real applications

Other papers investigate eigenvalue problems related to the vibrational properties of amorphous materials, ultrasound simulations on Xeon Phi accelerators, and parameter estimation for k-Wave simulations. The volume also explores analysis and visualization of HPC application behavior and graph connectedness for digital imagery. Together, these topics show how high-performance computing connects mathematical techniques with demanding computational workloads.

A proceedings volume for technical readers

As revised selected papers from an international conference, this book is suited to researchers, graduate students, and practitioners seeking research perspectives across high-performance and scientific computing. Its range makes it useful for readers exploring adjacent areas as well as those following a particular method or application.

For a focused snapshot of research presented at HPCSE 2019, this volume offers nine distinct contributions spanning computing systems, numerical analysis, and scientific applications.