为新兴的exascale架构实现具有性能可移植性的神经网络原子间模型
Implementing a neural network interatomic model with performance portability for emerging exascale architectures
摘要：
计算科学的两个主要推动力是更准确的预测和更快的计算;为此，分子动力学（MD）模拟的时代精神是追求机器学习和数据驱动的原子间模型，例如神经网络势，以及新颖的硬件架构，例如GPU。目前神经网络潜力的实现比传统的原子间模型慢了几个数量级，而迫在眉睫的exascale计算提供了使用这些模型运行大型，准确模拟的能力，使用新的和不同的exascale硬件实现MD的便携式性能需要重新思考传统算法，使用新的数据结构和库解决方案。我们在CabanaMD中重新实现了一个神经网络原子间模型，CabanaMD是一个MD代理应用程序，构建在为性能可移植性而开发的库上。我们的实现显示出显着改善的节点上的缩放在这个复杂的内核相比，目前的LAMMPS实现，在强和弱的缩放。我们的单源解决方案在许多情况下都能提高性能，其线程可扩展性可在单个CPU节点上模拟多达2100万个原子，在单个GPU上模拟多达200万个原子。我们还探索了并行性和数据布局选择（使用称为AoSoA的灵活数据结构）及其对性能的影响，通过选择正确的并行性和数据布局，GPU上的性能分别提高了约25%和约10%。
英文摘要：
The two main thrusts of computational science are more accurate predictions and faster calculations; to this end, the zeitgeist in molecular dynamics (MD) simulations is pursuing machine learned and data driven interatomic models, e.g. neural network potentials, and novel hardware architectures, e.g. GPUs. Current implementations of neural network potentials are orders of magnitude slower than traditional interatomic models and while looming exascale computing offers the ability to run large, accurate simulations with these models, achieving portable performance for MD with new and varied exascale hardware requires rethinking traditional algorithms, using novel data structures, and library solutions. We re-implement a neural network interatomic model in CabanaMD, an MD proxy application, built on libraries developed for performance portability. Our implementation shows significantly improved on-node scaling in this complex kernel as compared to a current LAMMPS implementation, across both strong and weak scaling. Our single-source solution results in improved performance in many cases, with thread-scalability enabling simulations up to 21 million atoms on a single CPU node and 2 million atoms on a single GPU. We also explore parallelism and data layout choices (using flexible data structures called AoSoAs) and their effect on performance, seeing up to ~25% and ~10% improvements in performance on a GPU simply by choosing the right level of parallelism and data layout, respectively.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
