AI 中文总结
Aether.jl是一款用Julia编写的高性能三维磁流体与多流体尘埃代码，采用交互式人机开发框架，单GPU性能优于C++代码，在Frontier超算4096个GCD上并行效率超93%，支持CPU、GPU运行，已公开。
AI 中文摘要
我们介绍Aether.jl，这是一个用Julia编写、主要面向GPU系统的可压缩流体动力学与磁流体动力学（MHD）有限体积新代码。该代码采用标准高阶Godunov方法，在笛卡尔、圆柱和球极坐标系中求解带约束传输的MHD方程，可通过刚性相互拖曳将任意数量的尘埃流体与气体耦合。它从零开始开发，采用交互式人机编码智能体工作流；本文档记录了该工作流框架及数值方法。性能关键内核通过KernelAbstractions编写，支持在多个厂商的CPU和GPU上运行。Aether.jl可通过交互式笔记本或批处理脚本运行，将原型开发、生产运行和分析整合在单一语言中。我们通过一系列流体动力学、MHD和尘埃测试验证了该实现。尽管用动态语言编写，Aether.jl在相同硬件上的单GPU吞吐量可达到甚至超过C++代码。在Frontier超算上的弱扩展测试中，4096个图形计算器件（GCD）的并行效率保持在93%以上。这些结果表明，动态语言现在可支持在 exascale 系统上进行生产级天体物理MHD模拟。Aether.jl及其Jupyter笔记本示例套件已公开可用。
英文摘要
We present $\texttt{Aether}$, a new finite-volume code for compressible hydrodynamics and magnetohydrodynamics, written in Julia and primarily designed for GPU systems. The code solves the MHD equations with constrained transport in Cartesian, cylindrical, and spherical-polar coordinates, using standard high-order Godunov methods. An arbitrary number of dust fluids can be coupled to the gas through stiff mutual drag. It was developed from scratch with interactive Human-coding agent workflow; the paper documents the framework of this workflow alongside the numerical methods. Performance-critical kernel is written through $\texttt{KernelAbstractions}$, and supports runs on CPUs and GPUs from multiple vendors. $\texttt{Aether}$ can be ran either from an interactive notebook or batch scripts, keeping prototyping, production runs, and analysis in a single language. We verify the implementation through a series of hydrodynamic, MHD, and dust tests. Although written in a dynamic language, $\texttt{Aether}$ achieves comparable or even higher single-GPU throughput than C++ code on the same hardware. In weak scaling on Frontier, parallel efficiency stays above $93\%$S on 4096 GCDs. These results show that a dynamic language now supports production astrophysical MHD simulations on exascale systems. $\texttt{Aether}$ and its Jupyter notebook example suite are publicly available.
Comments25 pages, 16 figures
