A new book for the shelves of supercomputing aficionados has emerged, which covers a rather impressive swath of topic areas—from analyses of current top systems like Sequoia and Titan, to a historical trend watch covering the last several years in high performance computing.
Contemporary High Performance Computing: From Petascale Toward Exascale, edited by Jeffrey S. Vetter at Oak Ridge National Lab, culls the thoughts of leading researchers and users at a number of major global supercomputing sites.
Among the highlights are contributions on HPC challenges in design and history from Jack Dongarra and Piotr Luszczek, an overview of the Green500 List from Wu Feng and Kirk Cameron, and other detailed chapters on contemporary systems, metrics and technologies from other recognizable thought leaders in high performance computing.
Side topics of exploration include HPC Clouds, as well as GPU computing and other accelerators, shared memory systems and more. The book is a total of 730 pages and is available in both hard copy ($99) and electronic versions—for some reason the e-version is still the same price. But that’s another conversation.
As the publishers describe:
The first part of the book examines significant trends in HPC systems, including computer architectures, applications, performance, and software. It discusses the growth from terascale to petascale computing and the influence of the TOP500 and Green500 lists. The second part of the book provides a comprehensive overview of 18 HPC ecosystems from around the world. Each chapter in this section describes programmatic motivation for HPC and their important applications; a flagship HPC system overview covering computer architecture, system software, programming systems, storage, visualization, and analytics support; and an overview of their data center/facility. The last part of the book addresses the role of clouds and grids in HPC, including chapters on the Magellan, FutureGrid, and LLGrid projects.
There aren’t many books available that seek to take stock of current systems and trends in supercomputing in such a comprehensive way, most likely because the space changes so frequently, causing a book to hit the outdated ranks in little time. However, there’s enough in the way of historical context for trends in accelerators, systems, applications and metrics that some degree of the material will have a lengthy shelf life.