A successful example of grid computing--Global PC Unite

A successful example of grid computing--Global PC Unite

Monday, June 2003 3:11 PM at the end of June 2000, IBM released the world's fastest supercomputer in the United States. The "ASCI White" computer, which performs 12.3 trillion operations per second, consists of 512 separate RS6000 servers designed in parallel, with a total of 8,192 processors running the IBM-developed dedicated UNIX system--aix. ASCI White is used to secure U.S. nuclear weapons at a price of $110 million. However, a previous SETI@home plan to search for extraterrestrial civilizations took advantage of the hundreds of thousands of PCs connected to it, gaining more processing power than any other supercomputer and faster computing than the ASCI.

May 1999 17, a science project to find extraterrestrial life signs at the University of California, Berkeley,--seti@home started, SETI@home is search for Extra terrestrial Intelligence at Home's abbreviation, meaning: Look for the alien civilization in the house. The SETI@home project mainly utilizes the idle capacity of networked PCs to analyze data obtained by the world's largest radio telescope to help scientists explore extraterrestrial life, and its computational model is essentially grid computing.

Since the official launch of the SETI@home project, 3 million volunteers have participated in the project, downloading snippets of information collected from radio telescopes from designated sites, running analysis with their own computers, looking for signs of life in the universe, and processing data up to 15T. On average, each participant made his or her computer work for SETI@home for 17.5 hours, which is equivalent to 482,023 years of working with a PC, which is equivalent to working with a supercomputer for 48 years. The project took full advantage of the power of computers distributed around the world, though the whole plan cost only 500,000 of dollars, but it has a powerful power.

The general process for SETI@home projects is this:

1. The Government or research department submits a task that requires a large amount of computation to the server in the form of procedures and data.

2. The server divides data and program code into smaller parts, also known as "subtasks."

The 3.PC machine installs a special client program that automatically contacts the server and automatically downloads and processes subtasks.

4. The results are returned to the server after the subtasks have been processed. The client then downloads the new subtasks and continues processing.

5. Once all of the subtasks have been processed, the server aggregates the results, generates the final report, and sends the final results back to the author.

SETI@home's information is derived from the information collected by the RICO Arecibo's radio telescope, which has collected about 35G of data a day since October 1998, since it does not have enough network bandwidth, so it can only record tapes and serve the greater Berkeley by traditional mail. The data is then divided into approximately 350K per unit, transmitted to the participants ' computers via the Internet, and distributed operations are used to process large amounts of data. Because the SETI@home project did not get the country's special appropriation fee, so had to ask the world's PC to help. The people involved in this event have no rewards, only silently dedication, but according to the feedback back to the number of data fragments statistics, from the start of the project to date, in the participation of the 202 members, the "return" of the top 10 in order: The United States, Britain, Canada, Germany, Japan, Sweden, Australia, Finland, the Netherlands, Norway, In Asia Japan and China Taiwan Province entered the top 20.

Although no aliens have been found so far, the SETI@home project has proved that grid computing is effective.