See the bottom of this page for a description of the icons on the page.

Note: This project is the new SETI@home, based on the BOINC platform Classic SETI@home is listed below. Classic SETI@home will be active for a while as users transition to the new SETI@home. Users are encouraged to migrate to the new SETI@home. User accounts have been migrated from Classic SETI@home to the new SETI@home, but users must activate them before they can use them. All user and team stats in the new SETI@home have been reset to 0. Read more about the transition.

See the BOINC platform information for the latest version of the BOINC client. Version 4.08 of the SETI@home application (which runs inside the BOINC client) is available for Windows as of November 19, 2004. Version 4.02 is available for Linux, Mac OSX and Solaris as of August 29, 2004.

See a guide for customizing the SETI@Home BOINC client graphics, and unsupported add-on tools available for the client. See information about porting and optimizing the BOINC SETI@home client. See a Powerpoint slide presentation about Berkeley Open Infrastructure for Network Computing (BOINC), the open-source software architecture used for the new SETI@home. These slides were used for a presentation at the 2002 O'Reilly Emerging Technology Conference.

See the status of the project servers.

View the new SETI@home discussion forum.

Note: This project is the Classic SETI@home. Users should participate in the new SETI@home project listed above. The new SETI@home project is active as of June 22, 2004. Classic SETI@home will be active for a while as users transition to the new SETI@home. User accounts have been migrated from Classic SETI@home to the new SETI@home, but users must activate them before they can use them. Read more about the transition.

Version 3.08 of the classic client is available for most platforms as of April 4, 2003. Version 3.08a is available for Mac OSX as of March 25, 2004: it "fixes screensaver problems under Mac OS 10.3.x." The latest newsletter, #22 ("SETI@home Reobservation Report"), was published April 28, 2004. The latest technical news report was published May 5, 2004. A Glossary of Concepts is available to explain the project's unique terminology. The project is also mapping the distribution of hydrogen in our galaxy. See the project's most promising signal candidates. See pictures of the Arecibo telescope from the SETI@Home team's March, 2003 visit. The project had its 5th birthday on May 18, 2004.

On April 4, 2003, a security hole was found in the SETI@home client and reported to the SETI@home team. Version 3.08 of the client, which fixes the security hole, is available for most platforms. On November 14, 2003, the project reported that one or more viruses/worms may be circulating on the Internet and installing SETI@home on infected computers. It has a page with more information.

SETI@Home is the largest public distributed computing project in terms of computing power: on September 26, 2001 it reached the ZettaFLOP (10²¹ floating point operations) mark--a new world record--performing calculations at an average of 71 TeraFLOPs/second. For comparison, the fastest individual computer in the world is IBM's ASCI White, which runs at 12.3 TeraFLOPs/second. On June 1, 2002, the project completed over 1 million CPU years of computation. On August 19, 2003, the project processed its 1 billionth work unit.

As of June 14, 2002, the project has found 3.2 billion spikes and 266 million Gaussians. See the best candidate signals found so far.

Check the status of the SETI@home data server and tapes.

View derived statistics for SETI@home.

Get answers to all of your SETI@home questions in the SETI@home FAQ.

View the Classic SETI@home discussion forum.

Read a short research paper, written in 1998, about the origins of SETI and SETI@home.

See a RealPlayer Video interview that Dr. David Anderson gave to CERN's GridCafe on April 30, 2004. The interview lasts 6.5 minutes. See an August 19, 2004, interview of Dan Werthimer, director of the SERENDIP SETI program and chief scientist of SETI@home, by Astroseti.org. Hear a December 6, 2004, interview of David Anderson by Planetary Radio.

On October 24, 2002, Laurence Loewe published the first scientific paper, "evolution@home: Experiences with work units that span more than 7 orders of magnitude in computational complexity," based on the results of the project. This paper was presented at CCGrid 2002 in May, 2002.

To participate, download the client, then select a run-file based on the amount of RAM you can dedicate to the application and the amount of time you want the application to run, then run the application and email the results file. See more information about the client in an overview. Release 6, for Windows and Macintosh, is available as of October 7, 2002. Please upgrade to this version if you are using an older version. Scheduling session 7 of the run-files is available as of March 27, 2003.

evolution@home GUI, a third-party tool, provides a graphical interface for the semi-automated version of the evolution@home client.

See high scores for the project.

Join an unofficial discussion forum about the project.

An old version of the site is available in German.

The application is built on the Fida and Mithral distributed computing platforms. The Windows client can run as a screensaver (which doesn't show any information about what it's doing), or as a command-line client (run the client.exe executable in the installed directory). Note that although you have the option to install the client in any directory, the screensaver expects it to be installed in C:\Program Files\UW Chemistry\ Eon (this bug will be fixed soon). The Linux client is command-line only. The application supports users behind proxy servers. Edit the client.cfg file: change active to yes, and add your proxy server hostname and port number. Note: on Windows, only use notepad to edit client.cfg. wordpad and the DOS edit command strip off important end-of-line characters in the file and client.exe will reset it to its default values. In dtpad you will see an empty rectangle character after each field: this is the end-of-line character and it must not be removed. http://www.free-dc.org/forum/forumdisplay.php?s=&forumid=59

Join an unofficial discussion forum, hosted by Free-DC, about the project.

The first climate models for a full 45-year beta-test simulation were successfully completed on March 6, 2003. The project received its 5,000th result on November 7, 2003. See some normal and abnormal results. In the 3 months after the project launch, it achieved: 9,796 completed full runs, 882,272 modelled years, 43,548 registered users, and the web site was translated into 14 languages. By April 6, 2004, the project completed 1.5 million years of simulation in over 22,000 runs. By July 5, 2004, the project completed 2 million years of simulation in over 30,000 runs. By October 18, 2004, the project completed 3 million years of simulation in over 40,000 runs. By December 14, 2004, the project completed 50,000 runs. The project began supporting a BOINC-based client on August 26, 2004. The project turned 1 on September 14, 2004: by that date "78,000 people in over 130 countries had completed 35,000 45-year GCM runs, computed 2.5 million model years and donated 6,000 years of computing time." On September 17, 2004, a book about using new technologies to sustain and protect natural ecosystems was released: chapter 12 of the book is about climateprediction.net and was written by several climateprediction.net team members.

The project clients have some large requirements. In particular, one work unit will take up to 6 weeks to complete on a 1.4 GHz CPU. Please study the requirements on the download page before downloading the clients, and please do not download the clients if you are not willing to commit to completing a work unit. The classic client runs as a graphical application (GUI) or as a service. The BOINC-based client is similar to other BOINC-based clients. See the BOINC platform information for the latest version of the BOINC client. The clients support users behind firewalls and proxy servers (see this page for details for the classic client). Users may run either client, but new features will only be added to the BOINC-based client. Version 2.2.28 of the classic client is available for Windows as of February 13, 2004: it adds some new features and fixes some bugs from previous versions. See the BOINC platform information for the latest version of the BOINC client. The BOINC client is available for Windows, Linux, and Mac OSX. Note: On February 4, 2004, the project announced that Microsoft Internet Explorer Security path (IE Update KB832894) prevents the classic climateprediction.net client from uploading or "trickling" results. A patch is being developed for that client.

An advanced visualization package is available for some Windows platforms as of July 15, 2004. Also, Windows users with Photoshop can download a Photoshop plug-in to make a 3D model of their simulation results.

Students and teachers can access school resources for the project.

Join a discussion forum for this project.

See technical reports and papers from this project.

The client does not need to contact a project server to get work. It submits results via ftp whenever it accumulates more than 100 Kbytes of results. The software also includes a separate ftp client which you can use to submit results manually. The Windows version of the client can be run as a screen-saver or from the command-line. Version 4.41f of the client is available for Windows as of October 18, 2004. Debugging version 4.41g of client is available as of November 6. This version should be tried by users of the 4.41f version if the 4.41f client crashes or stops. Version 4.33 and later can be run under Linux using Wine.

Join a discussion forum about the project.

Note: this project will end around January 1, 2005. Volunteers should choose a different project to participate in.

Results of the project will be used "for biodiversity research, education and conservation worldwide, especially to forecast environmental events and inform public policy with leading-edge science."

Anyone can access data and results from the project via the following services:

• Lifemapper Web Mapping Service - "a web service that allows clients to utilize the Lifemapper data without going through the Lifemapper website. With this service, a client can insert an image element into their own web page that will display a map alone or layered on top of other maps of the same area."
• Lifemapper QueryByLocation Service - "an XML web service which allows a client to query the two Lifemapper spatial databases to find the species present in a given area. The first database, DataPoints, consists of cached specimen locations from participating institutions. The second database, Models, consists of predicted habitat maps from our screen saver users. The area can be queried by a point (buffered 0), a circle (a point buffered up to .01 degrees), or a polygon."
• Lifemapper QueryTaxa Service - "an XML web service which allows a client to query the Lifemapper databases to find a species by ScientificName, by CommonName, or by Lifemapper TaxonID. Results can be limited to Lifemapper TaxonID, or can include Scientific Name, Genus, Species, and SubSpecies, DataProviders who provided data points for this species, number of Georeferenced Points and number of Models Computed."

To participate in the project, register for an account through the project, then download the graphical/screensaver or command-line client application. The client is available for Windows and Linux. The graphical version of the client "shows some information on the registered user that is running the application, what species is being processed at that time, a map displaying preliminary results found so far, as well as general information and messages downloaded from [the project] servers." The text version displays similar information in text messages: it runs somewhat faster because it doesn't have to display graphics. The text version of the client supports work unit caching, and keeps processing while downloading new work units or uploading results. The graphical version of the client will have this feature soon. The client supports users behind firewalls (it uses HTTP protocol and port 80), but does not support users behind proxy servers. Version 1.0.01 of the graphical client is available as of June 3, 2003. Version 1.0.03 of the command-line client is available for Windows as of February 26, 2004. Version 1.0.02 of the text client is available for beta testing on Linux as of November 19, 2003 and for beta testing on Mac OS X as of January 5, 2004.

Join an unofficial discussion group about this project.

From the project owner: "As an increasing number of mission critical tasks are automated, self-checking circuits are of paramount importance. For example in medical applications (heart monitors, pacemakers), transport (aeroplane hardware, traffic lights, car ABS braking), space (satellites, probes) and industrial facilites (nuclear power plants) and more to come in the future as cars start driving themselves, surgical operations are performed remotely, etc.. In all these areas human lives or great economic loss are at risk.

"The circuits produced by this projects are truly better than those of conventional design so would lead to safer controllers in all these applications saving lives and money."

On October 1, 2004, the project successfully evolved a large number "of circuits with full concurrent error detection using only 14% of the overhead required by the conventional approach," and began evolving a new generation of circuits "as big as those used in industry, many of them using hundreds of gates." On October 14, 2004, circuits and overhead figures from the project "were presented to experts in the self-checking field at the International On-Line Testing Symposium IOLTS 2004. One expert said DHEP may be the best method to design self-checking circuits."

The project uses Genetic Algorithms and Evolutionary Strategies to design improved circuits. Source code and documentation for the project are available for download.

To participate in the project, download and install the Java-based client and follow the directions on the website for configuring it for Linux/Unix or Windows. The client supports users behind firewalls or proxy servers, if those users have access to a SOCKS4 or SOCKS5 proxy: see setup instructions in the project website's FAQ/troubleshooting section. The client supports modem-users (your circuit population will only migrate when you are connected to the Internet, but will continue evolving whether or not you are connected).

See the project's current goal.

See the project's results so far.

Join a discussion forum about the project.

The project released its first status report on November 3, 2004. It successfully completed over 500,000 jobs by that date, with the help of 6,000 registered users and 7,500 active computers. It released its second status report on December 7, 2004. The project is officially shut down until the new project servers are available in early 2005.

The project uses a BOINC-based client, which runs an application called SixTrack. See the BOINC platform information for the latest version of the BOINC client. SixTrack's graphical screensaver displays a cross-section of the beam of particles that SixTrack is simulating. Version 4.47 of application is available for Windows as of October 26, 2004 and for Linux as of September 27, 2004. Linux users with NFS-mounted work directories should read the Known Bugs for the client.

Join a discussion forum about the project.