posted
As you may or may not know, your PC does not use 100% of it’s available resources all of the time. The aim of this email is for me to introduce you into the world of distributed computing. This is basically where you download a small unobtrusive program that will “crunch” data to help science based projects.
Crunching is the act of downloading work units (WU’s) from the project and computing the data and sending the results back. Also known as grid computing. The aim is to make the world’s largest supercomputer with lots of people’s home machines to help science related projects.
This guide will help you in setting up BOINC, the most dynamic of the available projects as it can run many projects. The program gives you credit for each WU you complete so that you can compare yourselves to others crunching each other BOINC project.
BOINC – This is the program used to download and crunch the data. This will be the user interface with the project. Download is small, only about 1.5Mb and once installed runs happily in your toolbar. This program time slices between multiple projects (more on that later). The BOINC projects are (with there web addresses):
BOINC Home Page - http://boinc.berkeley.edu/ Climate Prediction.net (CPDN) – http://climateprediction.net/ Search for Extraterrestrial Intelligence (SETI) – http://setiweb.ssl.berkeley.edu/ Large Hadron Collider (LHC) – http://athome.web.cern.ch/athome/ Einstein@home – http://einstein.phys.uwm.edu/ Predictor at home – http://predictor.scripps.edu/
Below is a brief outline of the pros and cons of each project and what it’s about science wise. The site will give you more information if you require it and will go into more detail than I do here..
CPDN : Climate predicting! Not much else t say! + Stable project, fairly instant credit system, Trickling credit system due to long WU time - Largest file size (600Mb sometimes…), takes around 400+ hours to complete one WU
SETI: The most crunched project, SETI searches radio telescope data to try and find a signal from aliens. + Good WU times (Take about 3 to 4 hours to crunch average WU) - Usually slow validation process (slow getting credit) and down quite often
LHC: Runs predictions and data from the Large Hadron Collider at CERN (Particle acceleration and destruction basically) + Short WU times - Down for upgrades at the minute, limited available WU’s
Predictor: Predicts and works out protein structures. + Very short WU times (about 1 hour) - Limited WU’s at the minute due to restart of project
Einstein: Looking for gravitational waves + Good credit per WU - In testing at the moment, CPU intensive
Thank you for reading this far. Hopefully you wish to help donate to science now and if you do, go to http://www.freewebs.com/tprcpdnstats/InstallingAndRunningBOINC/InstallingAndRunningBOINC.html for step by step instructions on installing and running BOINC.
If you want to know more about BOINC and distributed computing, feel free to go to http://forums.teamphoenixrising.net and ask any of us there.
Posts: 1 | Registered: Feb 2005
| IP: Logged |
Printer-friendly view of this topic
