PhD student at University of Science and Technology. Acquainted with HPC computation and omnipresent machine learning, currently focused on massively distributed simulation systems.
Efficient information flow in complex, often micro-scale simulation systems is a challenging task. In previous work we presented the way to utilize Akka in the process of creating highly concurrent, distributed, message-driven simulation system. Mechanisms of Akka Actors and Akka Clustering proven to be efficient enough to provide near linear scalability in range of tens of thousands of computational cores. We delved further into this topic to ensure that developed method does not influence correctness of simulation. Experimental tests show that wide variety of simulations can be implemented using our framework with negligible impact on the results.