| Grid Computing | | | | processors to achieve the same (or better) |
| Some consider this to be the "the third | | | | performance is less expensive. |
| information technology wave" after the Internet | | | | The future: during the past 10 years, the trends |
| and Web, and will be the backbone of the next | | | | indicated by ever faster networks, distributed |
| generation of services and applications that are | | | | systems, and multi-processor computer |
| going to further the research and development of | | | | architectures (even at the desktop level) clearly |
| GIS and related areas. | | | | show that parallelism is the future of computing. |
| | | | Distributed GIS |
| Parallel processing | | | | As the development of GIS sciences and |
| Parallel processing is the use of multiple CPU's to | | | | technologies go further, increasingly amount of |
| execute different sections of a program together. | | | | geospatial and non-spatial data are involved in |
| Remote sensing and surveying equipment have | | | | GISs due to more diverse data sources and |
| been providing vast amounts of spatial | | | | development of data collection technologies. GIS |
| information, and how to manage, process or | | | | data tend to be geographically and logically |
| dispose of this data have become major issues in | | | | distributed as well as GIS functions and services |
| the field of Geographic Information Science (GIS). | | | | do. Spatial analysis and Geocomputation are |
| To solve these problems there has been much | | | | getting more complex and computationally |
| research into the area of parallel processing of | | | | intensive. Sharing and collaboration among |
| GIS information. This involves the utilization of a | | | | geographically dispersed users with various |
| single computer with multiple processors or | | | | disciplines with various purposes are getting more |
| multiple computers that are connected over a | | | | necessary and common. A dynamic collaborative |
| network working on the same task. There are | | | | model " Middleware" is required for GIS application. |
| many different types of distributed computing, | | | | Computational Grid is introduced as a possible |
| two of the most common are clustering and grid | | | | solution for the next generation of GIS. Basically, |
| processing. | | | | the Grid computing concept is intended to enable |
| The primary reasons for using parallel computing | | | | coordinate resource sharing and problem solving in |
| are: | | | | dynamic, multi-organizational virtual organizations |
| Saves time. | | | | by linking computing resources with |
| Solve larger problems. | | | | high-performance networks. Grid computing |
| Provide concurrency (do multiple things at the | | | | technology represents a new approach to |
| same time). | | | | collaborative computing and problem solving in |
| Taking advantage of non-local resources - using | | | | data intensive and computationally intensive |
| available computing resources on a wide area | | | | environment and has the chance to satisfy all the |
| network, or even the Internet when local | | | | requirements of a distributed, high-performance |
| computing resources are scarce. | | | | and collaborative GIS. Some methodologies and |
| Cost savings - using multiple cheap computing | | | | Grid computing technologies as solutions of |
| resources instead of paying for time on a | | | | requirements and challenges are introduced to |
| supercomputer. | | | | enable this distributed, parallel, and high-throughput, |
| Overcoming memory constraints - single | | | | collaborative GIS application. |
| computers have very finite memory resources. | | | | Security |
| For large problems, using the memories of multiple | | | | Security issues in such a wide area distributed GIS |
| computers may overcome this obstacle. | | | | is critical, which includes authentication and |
| Limits to serial computing - both physical and | | | | authorization using community policies as well as |
| practical reasons pose significant constraints to | | | | allowing local control of resource. Grid Security |
| simply building ever faster serial computers. | | | | Infrastructure (GSI), combined with GridFTP |
| Limits to miniaturization - processor technology is | | | | protocol, makes sure that sharing and transfer of |
| allowing an increasing number of transistors to be | | | | geospatial data and Geoprocessing are secure in |
| placed on a chip. | | | | the Computational Grid environment. |
| However, even with molecular or atomic-level | | | | Conclusion |
| components, a limit will be reached on how small | | | | As the conclusion, Grid computing has the chance |
| components can be. | | | | to lead GIS into a new "Grid-enabled GIS" age in |
| Economic limitations - it is increasingly expensive | | | | terms of computing paradigm, resource sharing |
| to make a single processor faster. Using a larger | | | | pattern and online collaboration. |
| number of moderately fast commodity | | | | |