These terms seem to get thrown around rather vaguely, in my opinion, and was wondering if there were some hard-lined facts about what accounts for which category in these fields. If there aren't any, then, any heuristics for determining how to categorize a system would be helpful :)
closed as off topic by Artem Kaznatcheev♦, Tsuyoshi Ito, Neel Krishnaswami, Jeffε, Kaveh Nov 4 '11 at 15:20
Questions on Theoretical Computer Science Stack Exchange are expected to relate to research-level theoretical computer science within the scope defined by the community. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about reopening questions here. If this question can be reworded to fit the rules in the help center, please edit the question.
The grid computing paradigm emerged as a new field distinguished from traditional distributed computing because of its focus on large-scale resource sharing and innovative high-performance applications such as:
- distributed supercomputing;
- on demand computing;
- high throughput computing;
- data-intensive computing;
Distributed supercomputing applications require multiple supercomputers to solve problems otherwise too large or whose execution is divided on different components that can benefit from execution on different architectures. This class of applications present a number of challenges to be faced, like resource discovery and scheduling, coordinated startup, configuration at multiple sites, wide area message passing and fault tolerance. An example of such application is SF-Express, a distributed interactive simulation of a military battle. On demand computing refers to the possibility of dynamically acquiring online instruments (e.g. microscopes, satellite sensors and telescopes) connected by high-speed networks to gather and process the data generated. An experiment of microtomography at photon sources in 1999 demonstrated the feasibility of on demand computing. The aim of high throughput computing (HTC) is to schedule many independent jobs for parametric studies or data analysis; in this case a measure of efficiency is the number of jobs processed per unit of time. The two most important tools for HTC, namely Condor and Nimrod are now grid-aware. Data-intensive applications extract new knowledge from geographically distributed data archives or digital libraries; issues related to this class of applications include scheduling and configuration of multiple data flows through several hierarchy levels.
Finally cluster computing refers to the use of a cluster, which is a parallel computer, for running parallel scientific simulations. Therefore, when using just a single cluster you are doing parallel computing; when using more than one simultaneously but for the same application, may be with different architecture, you are doing distributed supercomputing, so this is grid computing. Platform computing, the company producing the LSF scheduler, promotes its LSF multi-cluster capabilities as grid computing.