General-purpose coordinator–master–worker model for efficient large-scale simulation over heterogeneous infrastructure

Abstract

In this work, we propose a general-purpose coordinator–master–worker (GP-CMW) model to enable efficient and scalable simulation. The model supports distributed and parallel simulation over a heterogeneous computing node architecture with both multi-core CPUs and GPUs. The model aims at maximizing the hardware activity rate while reducing the overall management overhead. The proposed model includes five components: coordinator, priority abstraction layer, master, hardware abstraction layer, and worker. The proposed model is mainly optimized for large-scale simulation that relies on massive parallelizable events. Extensive set of experimental results shows that GP-CMW provides a significant gain from medium to intensive simulation load by exploiting heterogeneous computing resources including CPU and GPU. Regarding simulation runtime, the proposed GP-CMW model delivers a speedup that is 3.6 times faster than the CMW model.

Keywords:

• heterogeneous computing
• large-scale simulation
• master–worker model
• GPGPU
• CUDA
• PADS

Statement of contribution

This paper proposes a general-purpose coordinator-master-worker (GP-CMW) simulation models to significantly increase the efficiency of large scale simulations when the number of events becomes large. GP-CMW operates on heterogeneous computing nodes by (a) managing the communication across multiple simulation instances through priority abstraction layer (PAL) to increase simulation stability, and (b) exploiting computing, data, and communication locality through hardware abstraction layer (HAL) to maximize simulation efficiency. Experimentation results shows that GP-CMW provides a significant gain from medium to intensive simulation load by exploiting heterogeneous computing resources including CPU and GPU. In terms of simulation runtime, it delivers a speedup of 3.6 times faster than basic CMW model.

Notes

1 www.ea.com/fr/sports