Gpu optimization accelerates quantum simulations of helium atom clusters
AI-Assisted GPU optimization of the Stochastic Variational Method for Few-Body Boson System
Distributed, Parallel, and Cluster Computing
Summary
Simulating tiny groups of helium atoms is important but very slow on regular computers. The authors created new computer code that uses graphics cards (GPUs) and artificial intelligence to speed up these simulations by more than 16 times on advanced GPUs. Their method batches many small calculations together and arranges data for efficient GPU use. This makes studying larger groups of atoms possible with much less waiting time.
What this means in practice
- •For quantum simulation engineers: Run faster quantum simulations of few-body atomic clusters by using the GPU-optimized SVM code for helium systems.
- •For high-performance computing developers: Adapt GPU batching and memory layout strategies from this work to improve throughput of small-matrix computations in scientific computing.
Authors
Noritaka Shimizu, Shigeyoshi Aoyama, Yusuke Tsunoda, Akira Nukada
Abstract
The stochastic variational method (SVM) is one of the most powerful methods to solve quantum few-body systems precisely in various fields, such as nuclear and atomic physics. To the best of our knowledge, no SVM code optimized for GPUs has been reported. We developed a parallel SVM code for few-body clusters of $^4$He atoms and optimized it for two GPU architectures, NVIDIA GH200 and AMD Instinct MI300A, with the entire code written by an AI coding agent. On the algorithmic side, we introduce the secular-equation method with the Gu--Eisenstat prescription for solving the generalized eigenvalue problem within the SVM framework. The main part of the code tuning is to batch the many small matrices so that the GPUs are used efficiently, together with an array layout in memory optimized for coalesced addressing. For the five-body system with 4800 SVM basis states, the tuned SVM code runs 16.2 (14.7) times faster on the NVIDIA GH200 (AMD MI300A) than the same tuned code on the Intel Xeon Max CPU system, and 168 (153) times faster than the first working CPU implementation. The achieved performance brings 6-body and larger cluster systems within reach.