HCDA

Yuhang Liu,Xian-He Sun,Yang Wang,Yungang Bao

DOI: https://doi.org/10.1145/3418291

IF: 22.7

2021-05-01

Communications of the ACM

Abstract:As a new era in computing emerges, so too must our fundamental thinking patterns.

computer science, theory & methods, software engineering, hardware & architecture

What problem does this paper attempt to address?

This paper attempts to address the issue of how to re-examine and expand Computational Thinking in the context of the new computing era to tackle the emerging challenges in computer system design. Specifically:

1. **The slowdown of Moore's Law and the rise of big data applications**: With the slowdown of Moore's Law and the increase in big data applications, traditional computational thinking is no longer sufficient to meet new challenges. The paper points out the need for a new, more general thinking paradigm to solve these problems.

2. **Memory access bottleneck**: Big data applications have made memory access a performance bottleneck, not just the performance of the computing core. Therefore, it is necessary to rethink computer system design from a data-centric perspective.

3. **The importance of historical thinking**: The paper emphasizes the importance of historical thinking in innovation. By reviewing past innovations and experiences, valuable references can be provided for future technological development.

4. **Multidimensional thinking paradigm**: The paper proposes a tetrahedral thinking paradigm that includes four fundamental thinking modes (Historical Thinking, Computational Thinking, Data-Centric Thinking, and Architectural Thinking). These thinking modes are interrelated and provide a unified framework for computer system design.

In summary, this paper aims to propose a new, more comprehensive thinking paradigm to address the new challenges in current computer system design, particularly the issues brought by memory access bottlenecks and big data applications.