QuietIce
Disabled
- Joined
- May 7, 2006
- Location
- Anywhere but there
What is SETI@Home?
SETI is the Search for Extra-Terrestrial Intelligence, which strives to answer the question "Are We alone in the Universe?". In 1999 The Planetary Society started SETI@Home as a screen saver, which utilized the idle time on home PCs to analyze data from the Arecibo, Puerto Rico, radio telescope. (The movie Contact was filmed in part on location at the observatory site.) SETI@Home is one of the first distributed computing projects and is certainly the largest, having had over one million users in countries all over the world during it's decade-long history. Currently there are almost 200,000 users and about 300,000 computers running SETI@Home. As of 2004, SETI@Home uses a general purpose distributed computing platform called BOINC, Berkley Open Infrastructure for Network Computing, developed at the University of California Berkeley. BOINC has a very user-friendly interface and is easy to set-up and maintain. It was designed specifically to run in the background with minimal interference on everyday computing.
How does SETI@Home work?
Berkeley has a collection of hardware that sits on top of the main receiver package at Arecibo. This hardware collects radio signals across many bands from whatever direction the telescope happens to be pointed. By using this 'piggy-back' method, SETI@Home does not directly use telescope time, it simply records what is already streaming into the telescope from space. These recordings are then sent to Berkeley where they are divided into blocks of data called 'work units'. The work units are cataloged and cued-up to be sent out to SETI@Home participants. The BOINC software, which regularly checks in with Berkeley, downloads work units as needed then analyzes the data for repeating signals and other anomalies, flagging certain results. When a work unit is complete the results are stored on the home computer until transfered to Berkeley, where the results are stored in a temporary database for future analysis. After transfer it's erased on the home machine to make room for more work units.
What happens if there's a computation error?
All work units are sent out to two different computers. When both results arrive at Berkeley they are compared to each other. If the results agree then both computers get credited for completing the work unit. If the results do not agree a third copy of the raw data is sent to a third computer to be processed and those results are compared to the first two results. Results are not stored in the science database until at least two computers have the same results, at which time those two computers are credited for completing the work unit. Bogus results are rare but they happen. By only recognizing duplicate results from different machines, Berkeley maintains the integrity of the data analysis process.
What are Cobblestones and RAC?
SETI@Home credit is officially awarded in units called "Cobblestones" for historic purposes but is often referred to as simply 'credits'. Each work unit is worth a certain number of Cobblestones, depending on the amount of processor time it takes to complete on a standardized computer system running the base SETI@Home application. Another term often seen in SETI@Home conversation is 'RAC', or Recent Average Credit. RAC is calculated using a fairly complex formula but is basically the average daily output of a given computer or participant over the last 15-20 days.
Who are the Overclockers.com SETI Team?
Overclockers.com currently ranks 6th in the world for most data processed. RAC varies over time but we have been moving back and forth between 5th and 7th in the world for the past couple of years. We feel this is very good because the number of members we have is fairly small compared to most other teams near us in the ranks. Part of the reason for this is overclocking. Since most of our computers are running 15-25% faster than their design speed we also process 15-25% more data than the average computer!
In addition there are SETI@Home 'optimizers' written by fellow SETI@Home participants. The base x86 Berkekely application is designed for use on any x86 computer (Intel and AMD machines most of us own) so it cannot take advantage of certain Instruction Sets, like SSE3 and SSSE4, imbedded in modern processors. Optimizers are specific applications that replace the standard Berkeley-issued application and are written specifically to take advantage of those advanced Instruction Sets. Berkeley is aware of these optimizers and although it cannot officially sanction them (because they would then be respensible for supporting them as well) they do recognise their use. In the end it's all about the science and if the optimizers speed up the data processing step no one's complaining.
How do I sign up for the Overclockers.com SETI Team?
Getting involved in SETI@Home and becoming an Overclockers.com SETI Team member is easy. All you need is a few minutes to set it up BOINC, a valid e-mail address, and a little idle computer time you're willing to donate to science. Just follow these step-by-step instructions.
It takes all of us to make the Team work.
Anything you can do to help is appriciated!
The "Useful information, FAQ and Instructions" thread started by Mictlan in 2003 can be found here.
SETI is the Search for Extra-Terrestrial Intelligence, which strives to answer the question "Are We alone in the Universe?". In 1999 The Planetary Society started SETI@Home as a screen saver, which utilized the idle time on home PCs to analyze data from the Arecibo, Puerto Rico, radio telescope. (The movie Contact was filmed in part on location at the observatory site.) SETI@Home is one of the first distributed computing projects and is certainly the largest, having had over one million users in countries all over the world during it's decade-long history. Currently there are almost 200,000 users and about 300,000 computers running SETI@Home. As of 2004, SETI@Home uses a general purpose distributed computing platform called BOINC, Berkley Open Infrastructure for Network Computing, developed at the University of California Berkeley. BOINC has a very user-friendly interface and is easy to set-up and maintain. It was designed specifically to run in the background with minimal interference on everyday computing.
How does SETI@Home work?
Berkeley has a collection of hardware that sits on top of the main receiver package at Arecibo. This hardware collects radio signals across many bands from whatever direction the telescope happens to be pointed. By using this 'piggy-back' method, SETI@Home does not directly use telescope time, it simply records what is already streaming into the telescope from space. These recordings are then sent to Berkeley where they are divided into blocks of data called 'work units'. The work units are cataloged and cued-up to be sent out to SETI@Home participants. The BOINC software, which regularly checks in with Berkeley, downloads work units as needed then analyzes the data for repeating signals and other anomalies, flagging certain results. When a work unit is complete the results are stored on the home computer until transfered to Berkeley, where the results are stored in a temporary database for future analysis. After transfer it's erased on the home machine to make room for more work units.
What happens if there's a computation error?
All work units are sent out to two different computers. When both results arrive at Berkeley they are compared to each other. If the results agree then both computers get credited for completing the work unit. If the results do not agree a third copy of the raw data is sent to a third computer to be processed and those results are compared to the first two results. Results are not stored in the science database until at least two computers have the same results, at which time those two computers are credited for completing the work unit. Bogus results are rare but they happen. By only recognizing duplicate results from different machines, Berkeley maintains the integrity of the data analysis process.
What are Cobblestones and RAC?
SETI@Home credit is officially awarded in units called "Cobblestones" for historic purposes but is often referred to as simply 'credits'. Each work unit is worth a certain number of Cobblestones, depending on the amount of processor time it takes to complete on a standardized computer system running the base SETI@Home application. Another term often seen in SETI@Home conversation is 'RAC', or Recent Average Credit. RAC is calculated using a fairly complex formula but is basically the average daily output of a given computer or participant over the last 15-20 days.
Who are the Overclockers.com SETI Team?
We run SETI@Home not just for the science but for the competition and companionship of a shared goal. Most of us run SETI part-time when we're using our computer for light tasks, like web browsing and e-mail, and/or when our computer is completely idle, like when we're away at work or school. Many of us have SETI 'farms'. Similar to server farms, SETI farms are a collection of two or more computers assembled for an express purpose - in our case processing as much SETI@Home data as possible in the shortest time possible. While a large part of our team's computing power does come from SETI farms an even larger part comes from all the part-time crunchers out there.Daniel Edgar, Founder, Overclockers.com SETI Team:
"We are a group of hobbyists who overclock our computers to achieve maximum performance."
Overclockers.com currently ranks 6th in the world for most data processed. RAC varies over time but we have been moving back and forth between 5th and 7th in the world for the past couple of years. We feel this is very good because the number of members we have is fairly small compared to most other teams near us in the ranks. Part of the reason for this is overclocking. Since most of our computers are running 15-25% faster than their design speed we also process 15-25% more data than the average computer!
In addition there are SETI@Home 'optimizers' written by fellow SETI@Home participants. The base x86 Berkekely application is designed for use on any x86 computer (Intel and AMD machines most of us own) so it cannot take advantage of certain Instruction Sets, like SSE3 and SSSE4, imbedded in modern processors. Optimizers are specific applications that replace the standard Berkeley-issued application and are written specifically to take advantage of those advanced Instruction Sets. Berkeley is aware of these optimizers and although it cannot officially sanction them (because they would then be respensible for supporting them as well) they do recognise their use. In the end it's all about the science and if the optimizers speed up the data processing step no one's complaining.
How do I sign up for the Overclockers.com SETI Team?
Getting involved in SETI@Home and becoming an Overclockers.com SETI Team member is easy. All you need is a few minutes to set it up BOINC, a valid e-mail address, and a little idle computer time you're willing to donate to science. Just follow these step-by-step instructions.
It takes all of us to make the Team work.
Anything you can do to help is appriciated!
The "Useful information, FAQ and Instructions" thread started by Mictlan in 2003 can be found here.
Last edited: