APICe » Theses » AlchemistTimeWarping

Parallelising Stochastic Simulation Algorithms: optimistic simulation engine for AlchemistĀ 


This thesis aims at exploring strengths and weaknesses of the optimistic execution in the context of the classical Stochastic Simulation Algorithms. This goal will be achieved by implementing a speculative parallel execution engine for Alchemist (featuring the Time Warping) and comparing it with a native high performance implementation. This thesis can lead to a valuable scientific contribution, if the author will be able to detect and describe some conditions in which the optimistic simulations should not run (e.g. some conditions about dependencies among reactions). It will be also possible to keep contact with Alessandro Pellegrini and its research team in La Sapienza, which currently works in the high performance optimistic simulation area.


Mirko Viroli (Supervisor)
Danilo Pianini (Co-supervisors)