Utilizing Volunteer Computing and Virtualization Technology for Climate Simulation
Date
2014-10-08
Authors
Liu, Kai
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The climatological community relies increasingly on computing intensive models and applications to study atmospheric chemistry, aerosols, carbon cycle and other tracer gases. These models and applications are becoming increasingly complex and bring geospatial computing challenges for scientists as follows: 1) enormous computational power is required for running these models and applications to produce results in a reasonable timeframe; 2) climate models are always sensitive and require special computing environments; 3) these models are challenging to provide convenient and fast solution to transfer the big data outputs from climate simulations. Presently, volunteer computing is getting more powerful and provides a potential solution for these problems by obtaining super computational resources from global volunteers. Meanwhile, the virtualization technology based on hardware or platforms allows researchers to run sensitive models in a predefined virtual machine. This thesis reports on research to integrate and optimize volunteer computing and virtualization technology for climate simulation based on the following: 1) utilizing volunteer computing resources so that the heterogeneous home computers can support climate applications; 2) utilizing virtualization technology to make the climate application run on different platforms; 3) optimizing the output collection mechanism to periodically upload climate model output; and 4) optimizing the credit system to grant credits periodically for long time climate simulation tasks. The research is based on NASA and George Mason University’s collaborative project Climate@Home, which is the first volunteer computing project using virtualization technology in the climate domain.
Description
Keywords
Volunteer computing, Virtualization technology, Climate stimulation