quTARANG: A high-performance computing Python package to study turbulence using the Gross-Pitaevskii equation

We present quTARANG, a robust GPU-accelerated Python package developed for a comprehensive study of turbulence problems in Bose-Einstein condensates (BECs). It solves the mean-field Gross-Pitaevskii equation (GPE) using a Time-splitting pseudo-spectral (TSSP) scheme and ground state calculations are performed using a Backward Euler spectral (BESP) scheme. quTARANG also has post-processing tools that can compute different statistical properties of turbulent Bose-Einstein condensates, such as kinetic energy spectra, particle number spectrum and corresponding fluxes. This paper provides detailed descriptions of the code, along with specific examples for calculating the ground state and turbulent state of the condensate under different initial conditions for both 2-D and 3-D cases. We also present results on the dynamics of the GPE in 2-D and 3-D used to validate our code. Finally, we compare the performance of quTARANG on different GPUs to its performance on a CPU, demonstrating the speedup achieved on various GPU architectures.本文提出quTARANG——一款专为玻色-爱因斯坦凝聚态（Bose-Einstein Condensates, BECs）湍流问题系统性研究开发的鲁棒型GPU加速Python软件包。该软件采用时间分裂伪谱（Time-splitting Pseudo-spectral, TSSP）格式求解平均场格罗斯-皮塔耶夫斯基方程（Gross-Pitaevskii Equation, GPE），基态计算则通过向后欧拉谱（Backward Euler Spectral, BESP）格式完成。quTARANG还集成了可计算湍流玻色-爱因斯坦凝聚态各类统计特性的后处理工具，例如动能谱、粒子数谱及其对应通量。本文详细阐述了该软件的实现细节，并针对二维与三维场景，给出了不同初始条件下凝聚态基态与湍流态计算的具体示例。此外，本文还提供了用于验证本软件正确性的二维及三维GPE动力学仿真结果。最后，我们对比了quTARANG在不同GPU与CPU上的运行性能，展示了各类GPU架构下获得的显著加速比。