Massively parallel fitting of Gaussian approximation potentials
Massively parallel fitting of Gaussian approximation potentials
复制标题
高斯近似势的大规模并行拟合
DOI：
10.1088/2632-2153/aca743
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Klawohn S
中科院分区：
文献类型：
--
作者：
Klawohn S
我们提出了一个数据并行的软件包，用于拟合高斯近似势（GAP）在多个节点上使用ScaLAPACK库与MPI和OpenMP。到目前为止，GAP模型的最大训练集大小受到单个计算节点上可用内存的限制。在我们的新实现中，描述符评估是并行进行的，没有通信要求。确定模型系数所需的后续线性求解与ScaLAPACK并行。我们的方法可扩展到数千个核心，解除了内存限制，并提供了大量的加速。这一发展将差距方法的适用性扩展到更复杂的系统，并为在更高级别的工作流程（如委员会模型或超参数优化）中有效嵌入GAP模型拟合提供了机会。
We present a data-parallel software package for fitting Gaussian approximation potentials (GAPs) on multiple nodes using the ScaLAPACK library with MPI and OpenMP. Until now the maximum training set size for GAP models has been limited by the available memory on a single compute node. In our new implementation, descriptor evaluation is carried out in parallel with no communication requirement. The subsequent linear solve required to determine the model coefficients is parallelised with ScaLAPACK. Our approach scales to thousands of cores, lifting the memory limitation and also delivering substantial speedups. This development expands the applicability of the GAP approach to more complex systems as well as opening up opportunities for efficiently embedding GAP model fitting within higher-level workflows such as committee models or hyperparameter optimisation.
影响因子：
9.7
作者：
Vandermause, Jonathan;Torrisi, Steven B.;Kozinsky, Boris
通讯作者：
Kozinsky, Boris
影响因子：
4.4
作者：
Imbalzano, Giulio;Zhuang, Yongbin;Ceriotti, Michele
通讯作者：
Ceriotti, Michele