TAMM：面向多体方法的张量代数框架
TAMM: Tensor Algebra for Many-body Methods
AI总结：
本文提出TAMM框架，通过解耦计算规格与执行，兼顾领域科学家算法开发和高性能计算优化，实现计算化学张量收缩方法的高效、可移植开发。
摘要：
在大规模计算平台上，计算化学中的张量收缩运算消耗了相当大比例的计算时间。由于在描述电子结构理论时广泛使用大型多维张量之间的张量收缩，这推动了多个面向异构计算平台的张量代数框架的发展。本文提出了面向多体方法的张量代数框架（TAMM），这是一个用于可扩展计算化学方法的高效开发且性能可移植的框架。TAMM框架将计算规格说明与这些运算在可用高性能计算系统上的执行解耦。基于这一设计选择，科学应用开发者（领域科学家）可以借助TAMM提供的张量代数接口专注于算法需求，而高性能计算开发者则可以专注于底层构造上的各种优化，例如高效的数据分布、优化的调度算法以及节点内资源（如GPU）的高效利用。TAMM的模块化结构使其能够扩展以支持不同的硬件架构并纳入新的算法进展。我们描述了TAMM框架以及我们在计算化学应用中可持续开发基于张量收缩方法的方式。我们展示了若干案例研究，突出其易用性以及与其他实现相比在性能和生产力方面的提升。
英文摘要：
Tensor contraction operations in computational chemistry consume significant fractions of computing time on large-scale computing platforms. The widespread use of tensor contractions between large multi-dimensional tensors in describing electronic structure theory has motivated the development of multiple tensor algebra frameworks targeting heterogeneous computing platforms. In this paper, we present Tensor Algebra for Many-body Methods (TAMM), a framework for productive and performance-portable development of scalable computational chemistry methods. The TAMM framework decouples the specification of the computation and the execution of these operations on available high-performance computing systems. With this design choice, the scientific application developers (domain scientists) can focus on the algorithmic requirements using the tensor algebra interface provided by TAMM whereas high-performance computing developers can focus on various optimizations on the underlying constructs such as efficient data distribution, optimized scheduling algorithms, efficient use of intra-node resources (e.g., GPUs). The modular structure of TAMM allows it to be extended to support different hardware architectures and incorporate new algorithmic advances. We describe the TAMM framework and our approach to sustainable development of tensor contraction-based methods in computational chemistry applications. We present case studies that highlight the ease of use as well as the performance and productivity gains compared to other implementations.
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。
