论文标题
86 pflops对1亿原子的深度分子动力学模拟，并具有从头开始准确性
86 PFLOPS Deep Potential Molecular Dynamics simulation of 100 million atoms with ab initio accuracy
论文作者
论文摘要
我们介绍了GPU版本的DEEPMD-KIT，该版本在训练深度神经网络模型时，可以以较高的精度来驱动非常大的分子动力学（MD）模拟。我们的测试表明，GPU版本的速度比具有相同功耗的CPU版本快7倍。代码可以扩展到整个峰会超级计算机。对于113、246、208原子的铜系统，该代码每天可以执行一个纳秒MD模拟，达到86个Pflops的峰值性能（占峰值的43％）。这种前所未有的能力以前精度进行了MD模拟的能力开辟了研究材料和分子中许多重要问题的可能性，例如异质催化，电化学细胞，辐照损伤，裂纹繁殖和生化反应。
We present the GPU version of DeePMD-kit, which, upon training a deep neural network model using ab initio data, can drive extremely large-scale molecular dynamics (MD) simulation with ab initio accuracy. Our tests show that the GPU version is 7 times faster than the CPU version with the same power consumption. The code can scale up to the entire Summit supercomputer. For a copper system of 113, 246, 208 atoms, the code can perform one nanosecond MD simulation per day, reaching a peak performance of 86 PFLOPS (43% of the peak). Such unprecedented ability to perform MD simulation with ab initio accuracy opens up the possibility of studying many important issues in materials and molecules, such as heterogeneous catalysis, electrochemical cells, irradiation damage, crack propagation, and biochemical reactions.
