Litcius/Paper detail

Simple heuristics for efficient parallel tensor contraction and quantum circuit simulation

Roman Schutski, Taras Khakhulin, Ivan Oseledets, Dmitry Kolmakov

2020Physical review. A/Physical review, A19 citationsDOIOpen Access PDF

Abstract

Tensor networks are the main building blocks in various computational sciences, ranging from many-body theory and quantum computing to probability and machine learning. Here we propose a parallel algorithm for the contraction of tensor networks using probabilistic graphical models. Given a tensor network, this algorithm splits the contraction task almost optimally into subtasks, which can be contracted independently in parallel. Our approach is based on the heuristic solution of the $\ensuremath{\mu}$-treewidth deletion problem in graph theory. We apply the resulting algorithm to the simulation of random quantum circuits and discuss the extensions for general tensor network contractions. According to our estimates, the lower bound on the simulation time of 1 million amplitudes of a 49-qubit circuit of depth 40 on the Summit supercomputer is $>5$ seconds.

Topics & Concepts

HeuristicsComputer scienceTensor contractionTheoretical computer scienceProbabilistic logicTensor decompositionTreewidthGraphical modelTensor (intrinsic definition)Quantum computerSimple (philosophy)Contraction (grammar)Graph theoryQuantumAlgorithmGraphMathematicsArtificial intelligenceTensor productLine graphCombinatoricsPhilosophyPure mathematicsMedicineInternal medicineEpistemologyOperating systemPhysicsPathwidthQuantum mechanicsQuantum many-body systemsQuantum Computing Algorithms and ArchitectureParallel Computing and Optimization Techniques