Einstein@home
Einstein @ Home ( D @ H ) is a project for distributed computing ( distributed computing ) of the LIGO Scientific Collaboration ( LSC). It locates in the Laser Interferometer Gravitational wave Observatory from the United States and the German GEO600 collected data for clues to gravitational waves of extremely dense, rapidly rotating compact stars. These include pulsars. According to the general theory of relativity, such stars deform spacetime around them, with measurable gravitational waves should arise. Since the end of March 2009 will also still looking binary radio pulsars in the data of Areciboteleskops.
Project Idea
The measurement of the gravitational wave is to be done by large detectors based on Michelson interferometers. The detectors should be accurate enough in the future to determine changes in length in the order of a fraction of the diameter of a proton. This involves large amounts of data that need even on today's supercomputers are based on cluster long time to complete analysis. In order to still be able cope with the analysis, they came up with the approach of distributed computing.
The project was officially launched on February 19, 2005 as part of the Year of Physics 2005. To manage the work packages, the BOINC framework is used. Significantly involved in the project is on German side, the Max Planck Institute for Gravitational Physics (Albert Einstein Institute) in Potsdam. There a big part of the scientific software is developed. The AEI is in March 2008 with its clusters Merlin ( 180 dual Athlon XP machines ) and Morgane (615 AMD Opteron nodes) is the second largest individual participants in the project. In a similar magnitude also provides the D-Grid initiative the project computational time. The others organized in the LSC research institutions provide the project hosts pools and cluster. Most of the work is, however, provided by the computers of now more than 528,000 volunteers, of which only about 57,000 participants are regularly active. In September 2010 the project were thus carried about 118,000 host computers than 300 teraflops of computing power available, which corresponded to the 14th place in the list of world's fastest supercomputers at this time. Long-term goal is the extraction of 400,000 regularly active players in order to analyze the data collected by the detectors in nearly real time.
Project History
Einstein @ Home began his analysis of 600 hours of data material from the LIGO S3 run, the accuracy of which, however, was still away a lot of the desired precision of the LIGO detector. The data had been previously examined for clusters, where no abnormalities were discovered. The first E @ H run with S3 data thus served mainly the test of scientific applications and a better calibration. Besides, numerous spurious signals were detected and removed. These interference signals come about by the sensitivity of the detectors. Especially by seismic disturbances, but also by signals from the power grid or the ocean surf hitting them from permanently. Of these disorders each detector is individually concerned. A gravitational wave would betray the fact that all detectors deflect simultaneously worldwide. According to the " cleansing" of the S3 data, this new version was analyzed again. In addition, some false signals were interspersed in order to make statements about the probability of detection of relevant signals between the disorders can. From the end of June 2005 to mid-2006 was the beginning of the analysis carried out in 2005 LIGO - run S4, which should reach an accuracy factor of 2. S5 should be the first pass that reaches the targeted accuracy. As part of the S5 LIGO detectors were operated continuously for a year. The analysis of the S5 data began in June 2006. S5R1 The first scan in this data set was completed in February 2007. There followed a brief search S5RI in a limited frequency range with a modified parameter set, which lasted until mid-April 2007. Meanwhile collected the detectors further data in the context of S5. The computational effort increases exponentially. To cope with the calculations, a new application has been tested in the search S5R2 that implements a hierarchical search. It will be initially only in a coarse grid, and later focused on the promising places. From 23 September 2007, the search began S5R3 with a second version of the algorithm for hierarchical search, which improves the sensitivity by about a factor of 6. The run S5R3b is a seamless continuation of the S5R3 in the frequency band above 800 Hz Since August 2008, the search S5R4a took place. Have already been carried out in 2008 test runs for the search for binary radio pulsars in December, 2009 Work Units for the Arecibo Binary Pulsar Search said application will be distributed to all participants of the Einstein @ Home project, unless the participants this is not clear in the user-specific settings since the end of March. After a few weeks of testing the GPU applications were released for Arecibo Binary Pulsar Search for Windows and Linux on 26 November 2009. While the majority of the calculations continue to take place on the CPU, the fast Fourier transforms are now calculated on the GPU, resulting at least for this part of the task to a considerable reduction of the computation time.
In June 2010, the project of the discovery of the pulsar PSR J2007 2722 hitherto unknown in the constellation fox succeeded. In March 2011, a second success was recorded with the discovery of the pulsar PSR J1952 2630 in the data from the year 2005 of the Arecibo Observatory. Until August 2012, 46 new pulsars were discovered by the project.
The project has exceeded the 1 petaflop barrier of calculation capabilities in January 2013 and is located on the computing power of the grid as seen on par with the 23 most powerful supercomputers in the world.
Projects
- Gravitational Wave S6 GC search: search for gravitational waves
- Binary Radio Pulsar Search ( Arecibo ): Search for binary pulsars in the data of the Arecibo telescope
- Gamma -ray pulsar search # 1: Search for gamma pulsars