ECLIPSE, a high-performance computing cluster, is currently achieving a significant milestone; it has surpassed 25 million computing hours. As one of the fundamental tools of theoretical research, ECLIPSE helps ELI Beamlines’ scientists with a detailed understanding of the processes involved in laser physics. The usage of the ECLIPSE cluster thus helps with the design of future experiments, scientific discoveries and their publications.
“The monthly price of a computing cluster service from a commercial provider with parameters similar to ECLIPSE would be $ 136 000. Twenty-five million computing hours would cost about $ 3 400 000. This is quite a decent amount of money – especially when compared to the cost of ECLIPSE, which was less than $ 600 000,” says Stefan Weber, head of the experimental program operating the cluster.
The main objective of the ECLIPSE cluster is to study the interaction of laser radiation with matter, in particular the fourth state of matter: plasma. Particle simulations that describe the relativistic motion of charged particles in the electric and magnetic fields of the laser and the plasma itself are most commonly used for this purpose. These simulations can be parallelized, i.e. executed on multiple computing servers at the same time, which makes the total simulation time considerably shorter. A typical simulation uses dozens of computing servers and takes about 1-2 days. Approximately 12 000 simulations are processed annually on the ECLIPSE cluster. Since the cluster currently has about sixty users, there is one task per user approximately every other day. Last but not least, the ECLIPSE computational cluster has contributed to the acquisition of many interesting scientific results, which are published in about a hundred scientific publications.
With advances in the research of laser interactions and due to the beginning of experimental research at ELI Beamlines, scientists are already reaching the technical limits of ECLIPSE. There is an increasing demand in terms of the number of simulations and their level of detail. These simulations help to understand experimental observations and to determine future research directions. Such increasing demands have led to the decision of significantly expanding ELI’s computing capabilities.