This edited book consists of 12 chapters, grouped into two parts and authored by a number of teams.
The first part, “Big Data Architectures,” includes four research chapters emphasizing state-of-the-art approaches to architectural aspects of high-performance computing (HPC). All four standalone chapters in the part introduce and discuss different state-of-the-art techniques within the field of hardware/software design optimization, parallel computing, and data flow issues in cloud computing.
Part 2, “Emerging Big Data Applications,” focuses more on applications of HPC in use, for example, state-of-the-art machine learning (ML) algorithms for handling big data applications such as genome projects. A particular focus is on the use of accelerators for performance improvement of the algorithms. In fact, the majority of state-of-the-art ML algorithms are computationally very expansive and in need of high-performance infrastructures. Accelerators such as field-programmable gate arrays (FPGAs) and graphics processing units (GPUs) have recently become more popular in solving computationally intensive algorithms and complex and large-scale problems. Although GPU technology in parallel computing has been under the radar for some time, their use for the purpose of accelerating ML algorithms still attracts the attention of researchers. On the other hand, FPGAs are subject to hardware/software design optimization for being well exploited, and also attract more and more attention for the very same purpose.
The authors mainly provide concise chapters with long lists of references, which can be very useful for junior researchers who need to understand the essence of the relevant field and extend understanding at the foundational level. The book could be more informative with extended and complementary chapters, which remains its main weakness. I would also like to note that the chapters in the second part are more related and complementary than those that make up the first part.