Information Technology (IT) is changing all aspects of life, including how scientific research is conducted. Over the years, a few grass-roots efforts developed in the geosciences that started to take advantage of these new resources. However, the impact of IT in the geosciences has not been at a high level yet. To improve understanding of IT and familiarize a group of students and researchers with these new tools, we have designed a Summer Institute for Geoscientists lecture series primarily focused on immediate needs of the community in the area of information technology. The goal of this Summer Institute is to educate a group of earth scientists and expand the community of IT users in earth science research.
the broad and rapid adoption of IT in science, and the advent
of major initiatives such as the NSF Cyberinfrastructure
program and UK’s E-science activity, it is essential
that the community at large be in a good position to take
full advantage of new opportunities to further improve research
and educational activities via IT-enabled approaches. This
Summer Institute is a step in this direction and designed
as an educational and outreach activity in the geosciences
primarily focusing on immediate needs. We have chosen a
format that provides a quick introduction to six key IT
topics that are relevant to the overall science goals of
the community in the immediate term. Each course will be
taught by an expert in that field and assisted by other
experts from the San Diego Supercomputer Center and NCAR.
Lectures will be followed with hands-on lab exercises. In
the following sections we describe each of these courses
Welcome & Introduction
(Instructors: Bertram Ludaescher, Kai Lin)
NOTE: Several of the sessions below include hands-on segments
Data Management (Continued)
Geographic Information Systems
Geographic Information Systems (Continued)
6:00PM Dinner Meeting
(Speaker: Chaitan Baru - "GEON")
(Instructor: Karan Bhatia)
At the high-level, Grid Computing consists of a methodology for creating "virtual organizations" (VO), that is, organizations that are composed of individuals and resources spread throughout other(non-virtual) organizations (NVO). Typically, the resources and individuals of the VO are geographically distributed, and resources are owned and managed by the NVO and donated for use in the VO subject to various security and use conditions. There are many technical challenges inherent in building a VO, including security, data management, resource discovery, user interfaces, application development. In this talk we address each of these issues, illustrating the current state-of-the-art and best practices within the grid community. In addition, as grid services merge with web services, its important to understand the similarities and differences. As such, the talk discusses the new services-based architecture for "grid services" and current efforts to make use of them.
(Instructors: Ilkay Altintas, Bertram Ludaescher)
(Instructor: Tim Kaiser)
We will start our talk with a definition of parallel high performance computing (HPC). This definition will lead us to a discussion of the advantages of HPC and then to the disadvantages of parallel computing.
There are many ways to categorize parallel computing. We will break the subject down into the shared memory and message passing paradigms. An overview will be given of each. We will then spend the remaining time discussing message passing parallel computing.
The primary library used for message passing is the Message Passing Interface (MPI). We will show simple examples of MPI based programs. This will lead us to the discussion of a simple scientific application. We will show how this application can be parallelized using MPI. Parallelization of this application demonstrates many of the common tasks in creating a real parallel program.
We will finish our discussion with an overview of some resources available for parallel programming.