# High Performance Computing for Geospatial Applications: A Retrospective View

**Authors:** Marc P. Armstrong  
**arXiv:** [1912.06548](https://arxiv.org/abs/1912.06548)  
**Date:** 2019-12-11  
**Summary page:** https://www.summarizepaper.com/en/paper/1912.06548v1/

## Key points

- High performance computing (HPC) is crucial for addressing the computational complexity of geospatial analyses
- Geospatial analyses often involve solution processes that require numerous iterations or combinatorial comparisons
- Rapidly growing volumes of geospatial information further complicate these analyses
- HPC helps in effectively handling these challenges
- Previous implementations of HPC in geospatial applications had limitations in scalability and portability
- Recent approaches leverage distributed resources like cyberinfrastructure, cloud computing, and fog computing for more flexibility and scalability
- The need for HPC in geospatial applications is emphasized due to computational complexity and increasing volume of geospatial data
- Utilizing distributed resources presents a promising avenue for efficient analysis and display of geospatial information.

## Summary

The chapter titled "High Performance Computing for Geospatial Applications: A Retrospective View" by Marc P. Armstrong discusses the importance of high performance computing (HPC) in addressing the computational complexity of geospatial analyses. These analyses often involve solution processes that require numerous iterations or combinatorial comparisons, making them computationally challenging. The author highlights that the complexity of geospatial analyses is further exacerbated by the rapidly growing volumes of geospatial information, which can impede analysis and display. Therefore, the application of HPC becomes crucial in effectively handling these challenges. The chapter provides an overview of past implementations of HPC in geospatial applications. It mentions that many previous applications were developed using hardware systems with short life-spans and implemented in software that was not easily portable. This limited their scalability and hindered their potential for wider adoption. However, recent approaches have shown promise by leveraging distributed resources such as cyberinfrastructure, cloud computing, and fog computing. These approaches offer more flexibility and scalability compared to traditional implementations. Overall, this chapter emphasizes the need for HPC in geospatial applications due to their computational complexity and the increasing volume of geospatial data. The shift towards utilizing distributed resources presents a more promising avenue for addressing these challenges and enabling efficient analysis and display of geospatial information.