Using Docker, An Industry Standard Technology To Run Gate Simulation On Multiple Platforms

Simulation is mandatory to initiate any new high energy physics or medical imaging experiments. These simulations generally consume a lot of computational resources. Parallelism on large clusters of computers is among the most efficient strategies to reduce simulation time [1] , [2] . These clusters are not easy to put in place since they are very expensive and power hungry. Furthermore, creating the environment to simulate on these clusters takes time because of the lack of high privilege access on the machines. Finally, because of the specificity of each computer grid, it is also difficult to share simulation environment and methodology between research groups. Nowadays companies like Amazon, Google, and Microsoft propose affordable cloud computing resources. In addition to those resources, containerization software like Docker enable the description of an environment in a simple text file called Docker image consequently easing the set-up and sharing of a simulation environment.