Scalable agent-based HPC Simulation of up to a million physics and signalling cells

After simulating a million cells on HPC environments, their final positions and simulation outputs are recorded in files which represent a dense cell spheroid. The simulation is a complex combination of the cell Physics and Signalling model which is distributed using MPI over a distributed memory network. The output of this data shows that the simulation generates reproducible results at scale.

在高性能计算（High Performance Computing, HPC）环境中完成百万级细胞模拟后，模拟所得的百万个细胞的最终位置与模拟输出结果被记录于存储致密细胞球体（cell spheroid）相关数据的文件中。本次模拟整合了细胞物理模型与信号传导模型，并通过消息传递接口（Message Passing Interface, MPI）在分布式内存网络上完成分布式部署与运行。该数据集的输出结果证实，本模拟可在大规模计算场景下生成可复现的实验结果。