Computer Architecture for Scientists: Principles and Performance

The dramatic increase in computer performance has been extraordinary, but it has key limits and structure. Software architects, developers, and data scientists need to understand how to exploit these concepts to harness them for future applications. Computer Architecture for Scientists covers four key pillars of computer performance and imparts a high-level basis for reasoning with and understanding these concepts.

Format: Hardback
Length: 264 pages
Publication date: 10 March 2022
Publisher: Cambridge University Press

The remarkable surge in computer performance has been truly extraordinary, yet it holds certain limitations and possesses a structured framework. Software architects, developers, and data scientists must grasp the underlying structure of computer performance to unlock its potential for future applications. Designed with upper-level undergraduates in mind, Computer Architecture for Scientists delves into four critical pillars of computer performance, offering a solid foundation for reasoning with and comprehending these concepts.

Small is fast: Understanding the impact of size scaling on performance is crucial. This principle highlights how reducing the size of data or instructions can lead to significant speed improvements. Implicit parallelism: Exploiting parallelism in sequential programs can significantly enhance their execution speed. By dividing tasks and utilizing multiple processing units, parallelism allows for efficient utilization of computational resources. Dynamic locality: Skirting physical limits by arranging data in a smaller space is a key strategy for optimizing performance. This involves techniques such as cache management and memory optimization to minimize data access times and improve overall system efficiency. Parallelism: Increasing performance through the collaboration of teams of workers is a fundamental principle of computer architecture. By leveraging parallelism, tasks can be divided and executed simultaneously, leading to faster and more efficient computations.

These principles and models provide accessible high-level insights and quantitative modeling, avoiding excessive low-level detail that may distract from the core concepts. Furthermore, the text delves into the growing importance of GPU (Graphics Processing Unit) and machine-learning accelerators in mainstream applications. These specialized devices have revolutionized the computing landscape by enabling high-performance computing and advanced artificial intelligence tasks.

In conclusion, the dramatic increase in computer performance has opened up a world of possibilities for innovation and advancement. By understanding the fundamental structure of computer performance and leveraging its principles, software architects, developers, and data scientists can create powerful applications that push the boundaries of what is possible. The integration of GPU and machine-learning accelerators has further expanded the horizons of computing, enabling the realization of complex tasks and the development of new technologies. As we continue to explore the potential of computer architecture, we can expect even more remarkable developments in the years to come.

Weight: 644g
Dimension: 176 x 250 x 19 (mm)
ISBN-13: 9781316518533
Edition number: New ed