17-02-2011, 03:43 PM
[attachment=8750]
Grid Computing
Scenario
For years, Dr. Rayburn has been looking for tools to helphis architecture students move beyond paper sketchesand scaled-down models. He knows that as workingarchitects, they will be using computer simulations thatrequire not just design skill but proficiency with increasinglycomplex software and hardware. Unfortunately,his department cannot afford to purchase and supporta computing system with the necessary processing capacityto run such advanced applications.Over the summer, the university’s IT staff, working withthe computer science department, set up a computergrid running on the campus network. The grid connectsnearly all university-owned computers, includingthose in labs, the library, as well as faculty and staffoffices. The software that runs the grid gives local userspriority for those machines, but when they are idle,their processors can be used over the grid. Using thepower of the campus grid, Dr. Rayburn’s students cannow use sophisticated architectural design softwarethat previously was unavailable because of its processingrequirements. With the software, students candesign buildings and other structures as well as the areassurrounding them, and create three-dimensional,interactive animations of their designs. As presentations,the animations allow viewers to “fly” over andaround the scenes the students generate, zooming inand out and moving in any direction they want to go.The university’s grid supplies enough unused computingpower to process the animations fast enough for itall to function smoothly.After several weeks of using the software, two of Dr.Rayburn’s students persuade faculty in the meteorologydepartment to connect a very large climatic databaseto the grid. The database includes data about theexact positioning of the sun and moon at any latitudeon the globe during daily, monthly, and yearly cycles,as well as historical data on weather conditions formost parts of the world. With the database availableon the grid, the students can incorporate seasonalchanges into their animations. They can render a buildingat a particular latitude, at a specific time of the yearor spanning weeks or months. Dr. Rayburn sees thatwith the new capabilities, his students are able to createbetter designs, ones that make more creative useof natural light—even as seasons change—and thatdemonstrate students’ deliberation about how theirstructures interact with the environment.
What is it?
Computing grids are conceptually not unlike electrical grids. In anelectrical grid, wall outlets allows us to page link to an infrastructure ofresources that generate, distribute, and bill for electricity. When youconnect to the electrical grid, you don’t need to know where thepower plant is or how the current gets to you. Grid computing usesmiddleware to coordinate disparate IT resources across a network,allowing them to function as a virtual whole. The goal of a computinggrid, like that of the electrical grid, is to provide users withaccess to the resources they need, when they need them.Grids address two distinct but related goals: providing remoteaccess to IT assets, and aggregating processing power. The mostobvious resource included in a grid is a processor, but grids alsoencompass sensors, data-storage systems, applications, andother resources. One of the first commonly known grid initiativeswas the SETI@home project, which solicited several million volunteersto download a screensaver that used idle processor capacityto analyze data in the search for extraterrestrial life. In a morerecent example, the Telescience Project provides remote access toan extremely powerful electron microscope at the National Centerfor Microscopy and Imaging Research in San Diego. Users of thegrid can remotely operate the microscope, allowing new levels ofaccess to the instrument and its capabilities.
Who’s doing it?
Many grids are appearing in the sciences, in fields such as chemistry,physics, and genetics, and cryptologists and mathematicianshave also begun working with grid computing. Grid technology hasthe potential to significantly impact other areas of study with heavycomputational requirements, such as urban planning. Anotherimportant area for the technology is animation, which requiresmassive amounts of computational power and is a common tool ina growing number of disciplines. By making resources available tostudents, these communities are able to effectively model authenticdisciplinary practices.
How does it work?
Grids use a layer of middleware to communicate with and manipulateheterogeneous hardware and data sets. In some fields—astronomy, for example—hardware cannot reasonably be movedand is prohibitively expensive to replicate on other sites. In other
