Computers in the design and analysis of HEP experiments

Abstract

Experimentalists in elementary particle physics have always been among the world's biggest consumers of computing resources, and techniques have now developed to the point where nearly every aspect of our science, from the design of the accelerators to the publishing of the physics results, makes such extensive use of computers that they have become truly indispensable. We describe the more important computing techniques used for the biggest experiments of both the current generation (CERN's SPS and LEP, and the Fermilab collider) and the next generation (LHC and SSC) of particle accelerators and storage rings. First we cover some of the Software Engineering methods needed to allow hundreds of physicists to work together to produce and maintain the hundreds of thousands of lines of computer code required for the different components of software for the big experiments: (1) the data acquisition and detector control, (2) the event and detector simulation, and (3) the reconstruction and analysis of the data. Then we will see how the experimentalist visualizes either real or simulated events, and how he can analyze Gigabytes of data interactively. To acquire the required reliability and accuracy in the simulation requires new techniques for pseudorandom number generation and we are studying the possibility of using quasi-Monte Carlo techniques.