Multi-agent based simulation has been widely used in behavior finance, and several single-processed simulation platforms with Agent-Based Modeling (ABM) have been proposed. However, traditional simulations of stock markets on single processed computers are limited by the computing capability since financial researchers need larger and larger number of agents and more and more rounds to evolve agents' intelligence and get more efficient data. This paper introduces a distributed multi-agent simulation platform, named PSSPAM, for stock market simulation focusing on large scale of parallel agents, communication system and simulation scheduling. A logical architecture for distributed artificial stock market simulation is proposed, containing four loosely coupled modules: agent module, market module, communication system and user interface. With the customizable trading strategies inside, agents are deployed to multiple computing nodes. Agents exchange messages with each other and with the market based on a customizable network topology through a uniform communication system. With a large number of agent threads, the round scheduling strategy is used during the simulation, and a worker pool is applied in the market module. Financial researchers can design their own financial models and run the simulation through the user interface, without caring about the complexity of parallelization and related problems. Two groups of experiments are conducted, one with internal communication between agents and the other without communication between agents, to verify PSSPAM to be compatible with the data from Euronext-NYSE. And the platform shows fair scalability and performance under different parallelism configurations.

In parallel computing systems using the master/worker model for distributed grid computing, as the size of handling data grows, the increase in the data transmission time degrades the performance. For divisible workload applications, therefore, multiple-round scheduling algorithms have been being developed to mitigate the adverse effect of longer data transmission time by dividing the data into chunks to be sent out in multiple rounds, thus overlapping the times required for computation and transmission. However, a standard multiple-round scheduling algorithm, Uniform Multi-Round (UMR), adopts a sequential transmission model where the master communicates with one worker at a time, thus the transmission capacity of the link attached to the master cannot be fully utilized due to the limits of worker-side capacity. In the present study, a Parallel Transferable Uniform Multi-Round algorithm (PTUMR) is proposed. It efficiently utilizes the data transmission capacity of network links by allowing chunks to be transmitted in parallel to workers. This algorithm divides workers into groups in a way that fully uses the link bandwidth of the master under some constraints and considers each group of workers as one virtual worker. In particular, introducing a Grouping Threshold effectively deals with very heterogeneous workers in both data transmission and computation capacities. Then, the master schedules sequential data transmissions to the virtual workers in an optimal way like in UMR. The performance evaluations show that the proposed algorithm achieves significantly shorter turnaround times (i.e., makespan) compared with UMR regardless of heterogeneity of workers, which are close to the theoretical lower limits.