Parallelisation Sample Clauses

Parallelisation. VirtU will provide access to the wealth of simulation data contained within the TVO, at the simplest level giving access to the raw data in a custom-designed format, but also giving access to refined data products such as group catalogues, merger histories and semi-analytic galaxy populations. To help the TVO user with their research, we plan to provide a suite of basic software tools which are used commonly by theorists. These include group finders (such as the code used to make merger trees), and codes for evaluating power spectra, correlation functions of objects sets which may be, for example, simulation particles or semi-analytic galaxies. A number of efficient algorithms for these relatively simple kinds of analyses have been developed over the last few decades. However, most extant codes in use are serial codes. In order for the TVO to be able to handle large datasets it is important to parallelise these codes. We therefore propose in the Project Plan some effort on parallelising the analyses tools. In the case of the group finder, this work is required not only to provide a tool for the user, but also to generate the group catalogues and the merger trees which are fed into the semi-analytic codes in order to generate the prime content of the TVO.