速度的形状:分布式量子电路模拟中分区几何形状与秩密度的影响
The Shape of Speed: Impacts of Partition Geometry and Rank Density in Distributed Quantum Circuit Simulations
浏览论文内容
中文总结 AI 辅助
该研究通过大规模实验证明,分布式量子电路模拟中分区几何形状主导性能,近立方分区比扁平分区快1.73-2.31倍,且每节点16秩可进一步减半求解时间并降低能耗。
中文摘要 AI 辅助
在分布式量子电路模拟中,形状不佳的分区可能在计算开始前就将性能减半。在 Fugaku 上对 764 个经过验证的配置(涵盖十二种算法、十三种三维环面分区几何形状以及六种秩密度,用于在 1,024 个节点上进行 39 量子比特模拟)进行的评估表明,分区几何形状主导运行时间。所有十二种算法在扁平分区上的运行速度比在近立方分区上慢 1.73-2.31 倍,尽管数据传输量相同,这证明性能下降源于网络传输而非通信量。该惩罚随三维环面分区的纵横比缩放(运行时间 $\propto a^{0.39}$,$r = 0.72$)。秩密度是次要因素,仅在紧凑几何形状下,每节点 16 个秩可将运行时间缩短 11%。最终,相对于扁平分区,请求每节点 16 个秩的近立方分区可将求解时间大致减半,同时扁平分区还多消耗 1.82 倍的能量。一项无模拟器的全对全微基准测试证实,对于以集合通信为主的工作负载,存在类似的几何形状惩罚。
英文摘要
In distributed quantum circuit simulation, a poorly shaped partition can halve performance before computation begins. Evaluation on Fugaku across 764 validated configurations (twelve algorithms, thirteen torus partition geometries, and six rank densities for 39-qubit simulations on 1,024 nodes) shows that partition geometry dominates runtime. All twelve algorithms run 1.73-2.31x slower on flat partitions than on near-cubic ones despite identical data transfer, proving the slowdown stems from network delivery rather than communication volume. This penalty scales with the 3D torus partition aspect ratio (runtime $\propto a^{0.39}$, $r = 0.72$). Rank density is secondary, cutting runtime by 11% at 16 ranks per node only on compact geometries. Ultimately, requesting a near-cubic partition with 16 ranks per node roughly halves time-to-solution relative to flat partitions, which also consume 1.82x more energy. A simulator-free all-to-all microbenchmark confirms a similar geometry penalty for collective-dominated workloads.
发表机构
- RIKEN Center for Computational Science(RIKEN计算科学中心)
- Keio University(庆应义塾大学)
机构由 AI 辅助整理,请以论文原文为准。