Data Management: Interactions with Computer Architecture and Systems

by Xiaodong Zhang, Rubao Lee

Overview

This guide illuminates the intricate relationship between data management, computer architecture, and system software. It traces the evolution of computing to today's data-centric focus and underscores the importance of hardware-software co-design in achieving efficient data processing systems with high throughput and low latency. The thorough coverage includes topics such as logical data formats, memory architecture, GPU programming, and the innovative use of ray tracing in computational tasks. Special emphasis is placed on minimizing data movement within memory hierarchies and optimizing data storage and retrieval. Tailored for professionals and students in computer science, this book combines theoretical foundations with practical applications, making it an indispensable resource for anyone wanting to master the synergies between data management and computing infrastructure.

Product Details

ISBN-13:	9781009123310
Publisher:	Cambridge University Press
Publication date:	11/28/2024
Pages:	304
Product dimensions:	6.97(w) x 9.80(h) x 0.91(d)

About the Author

Xiaodong Zhang is Robert M. Critchfield Professor in Engineering and University Distinguished Scholar at the Ohio State University. He specializes in data management in computer and distributed systems. His influential research is broadly adopted in various sectors. Notably, he received the ACM MICRO Test of Time Award in 2020 and is a Fellow of both ACM and IEEE.

Rubao Lee is a distinguished computer scientist who has made significant contributions to GPU-accelerated database systems and data processing. His innovations, like RCFile and YSmart, are widely adopted in industry. He received ICDCS's 2011 Best Paper Award and the Ohio State University's 2018 Lumley Research Award.

Table of Contents

About the authors; Preface; 1. Introduction; 2. Data storage: physical allocation and logical format; 3. Main memory: the physical space; 4. Buffer replacement algorithms; 5. In-memory data processing in large data centers; 6. GPU computing: a new algorithm-to-architecture interaction; 7. GPU for structured data; 8. GPU for spatial data: a case study in pathology imaging applications; 9. Ray tracing hardware in GPUs for accelerated computation; 10. The future of computing – synergies in data management and system architecture; Bibliography; Index.

Editorial Reviews

‘Data Management: Interactions with Computer Architecture and Systems is a must-have for any software engineer. This comprehensive book distills decades of research and engineering practices into a clear and insightful guide. Readers will gain a systematic understanding of cutting-edge advancements, making it an invaluable resource for solving real-world problems.’ Kaibo Wang, Uber Technologies

‘This textbook is a timely and comprehensive resource that explores the strong interactions among data management, memory hierarchy, computation, and applications. It strikes an elegant balance between fundamental knowledge and advanced topics, illustrating how data-centric algorithms, architectural and system designs, and software stacks should work together to achieve the high efficiency critically needed in the modern era of computing. This textbook is a natural sequel to conventional architectural books, designed for data applications ranging from computer graphics and databases to machine learning. It is far-reaching and future-proof!’ Deming Chen, University of Illinois Urbana-Champaign

‘This book breaks down the barriers between two fundamental computer science subjects that have been always taught separately, database systems and computer architectures. The book is a great read needed to see the full picture and the interplay between these two important subjects. Highly recommended for those seeking the understanding of system internal decisions and how do they impact various system components.’ Mohamed Mokbel, University of Minnesota

‘The hardware platforms on which database systems are developed have changed drastically since the early days of databases. The last 2+ decades have seen the research and development community study the impact of these changes on the database system architecture and internals and significant developments have occurred. The relationship between database systems and operating systems have been a concern since at least 1880s. We have not had a coherent book that covers the complex interactions between database systems, operating systems and the underlying hardware platform. This book by Zhang and Lee provides an excellent coverage of this area. I enjoyed reading it and I expect it to be very useful in our teaching.’ M. Tamer Özsu, University of Waterloo

From the Publisher