Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2111.10466

Simulation of quantum physics with Tensor Processing Units: brute-force computation of ground states and time evolution

Markus Hauru, Alan Morningstar, Jackson Beall, Martin Ganahl, Adam G. M. Lewis, Guifré Vidal

Published Nov 19, 2021• 5 citations
arXivHTMLPDF

Abstract

Tensor Processing Units (TPUs) were developed by Google exclusively to support large-scale machine learning tasks. TPUs can, however, also be used to accelerate and scale up other computationally demanding tasks. In this paper we repurpose TPUs for the challenging problem of simulating quantum spin systems. Consider a lattice model made of $N$ spin-$\frac{1}{2}$ quantum spins, or qubits, with a Hamiltonian $H = \sum_i h_i$ that is a sum of local terms $h_i$ and a wavefunction $|Ψ\rangle$ consisting of $2^N$ complex amplitudes. We demonstrate the usage of TPUs for both (i) computing the ground state $|Ψ_{gs}\rangle$ of the Hamiltonian $H$, and (ii) simulating the time evolution $|Ψ(t)\rangle=e^{-itH}|Ψ(0)\rangle$ generated by this Hamiltonian starting from some initial state $|Ψ(0)\rangle$. The bottleneck of the above tasks is computing the product $H |Ψ\rangle$, which can be implemented with remarkable efficiency utilising the native capabilities of TPUs. With a TPU v3 pod, with 2048 cores, we simulate wavefunctions $|Ψ\rangle$ of up to $N=38$ qubits. The dedicated matrix multiplication units (MXUs), the high bandwidth memory (HBM) on each core, and the fast inter-core interconnects (ICIs) together provide performance far beyond the capabilities of general purpose processors.

DOI
10.48550/arxiv.2111.10466

Related papers

Simulation of Quantum Many-Body Dynamics with Tensor Processing Units: Floquet Prethermalization

Alan Morningstar, Markus Hauru, Jackson Beall +4 more

Tensor processing units (TPUs) are specialized hardware accelerators developed by Google to support large-scale machine-learning tasks but they can also be leveraged to accelerate and scale other linearalgebra-intensive computations.In this paper, we demonstrate the usage of TPUs for massively parallel classical simulations of quantum many-body dynamics on long time scales.We apply our methods to study the phenomenon of Floquet prethermalization, i.e., exponentially slow heating in quantum spin chains subject to high-frequency periodic driving.We simulate the dynamics of L = 34 qubits for over 10 5 Floquet periods, corresponding to circuits with 4 × 10 6 nearest-neighbor two-qubit gates.The circuits simulated have no additional symmetries and represent a pure-state evolution in the full 2 L -dimensional Hilbert space.This is achieved by distributing the computation over 128 TPU cores.On that size TPU cluster, we find speed-ups in wall-clock run time of 230 times and 15 times when compared to reference CPU and single-graphics-processing-unit (GPU) simulations, respectively, for shorter-time 30-qubit simulations that can be handled by all three platforms.We study the computational cost of the simulations, as a function of both the number of qubits and the number of TPU cores used, up to our maximum capacity of L = 40 qubits, which requires a "full pod" of 2048 TPU cores with tens of terabytes of memory in total.For these simulations, an eight-TPU-core machine is comparable to a single A100 GPU and thus the full TPU pod is comparable to a machine with hundreds of top-of-the-line GPUs.However, the TPU pod is more energy and cost efficient and readily accessible (via Google Cloud), unlike such large many-GPU configurations.We also study the accumulation of numerical error as a function of circuit depth in very deep circuits.Our work demonstrates that TPUs can offer significant advantages for state-of-the-art simulations of quantum many-body dynamics.

arXiv:2111.08044•PRX Quantum•May 11, 2022•30 citations

Faster quantum state decomposition with Tucker tensor approximation

Stanislav Protasov, Marina Lisnichenko

Abstract Researchers have put a lot of effort into reducing the gap between current quantum processing units (QPU) capabilities and their potential supremacy.One approach is to keep supplementary computations in the CPU, and use QPU only for the core of the problem. In this work, we address the complexity of quantum algorithm of arbitrary quantum state initialization, an important building block of quantum data analysis and machine learning.QPUs do not outperform classical machines with existing precise initialization algorithms.Hence, many studies propose an approximate but robust quantum state initialization.Cutting a quantum state into a product of (almost) independent partitions with the help of CPU reduces the number of two-qubit gates, and correspondingly minimizes the loss of state fidelity in the quantum part of the algorithm.To find the least entangled qubits, current methods compute the singular value decomposition (SVD) for each qubit separately with CPU.In this paper, we optimize CPU usage and memory resource bottlenecks.We consider Tucker tensor decomposition as an alternative to the CPU-based SVD in a single low-entangled qubit detection task without the loss of solution quality.Both proposed methods outperform the SVD in time and memory for systems of at least ten qubits. We achieve an order faster implementation and two orders less memory usage for a system of 15 qubits.

Quantum Machine Intelligence•Jul 4, 2023•2 citations

Tensor Processing Units for Financial Monte Carlo

Francois Belletti, Davis King, Kun Yang +4 more

Monte Carlo methods are critical to many routines in quantitative finance such as derivatives pricing, hedging and risk metrics. Unfortunately, Monte Carlo methods are very computationally expensive when it comes to running simulations in high-dimensional state spaces where they are still a method of choice in the financial industry. Recently, Tensor Processing Units (TPUs) have provided considerable speedups and decreased the cost of running Stochastic Gradient Descent (SGD) in Deep Learning. After highlighting computational similarities between training neural networks with SGD and simulating stochastic processes, we ask in the present paper whether TPUs are accurate, fast and simple enough to use for financial Monte Carlo. Through a theoretical reminder of the key properties of such methods and thorough empirical experiments we examine the fitness of TPUs for option pricing, hedging and risk metrics computation. In particular we demonstrate that, in spite of the use of mixed precision, TPUs still provide accurate estimators which are fast to compute when compared to GPUs. We also show that the Tensorflow programming model for TPUs is elegant, expressive and simplifies automated differentiation.

Society for Industrial and Applied Mathematics eBooks•Jan 1, 2020•20 citations

Tensor Processing Units for Financial Monte Carlo

Francois Belletti, Davis King, Yang, Kun +4 more

Monte Carlo methods are critical to many routines in quantitative finance such as derivatives pricing, hedging and risk metrics. Unfortunately, Monte Carlo methods are very computationally expensive when it comes to running simulations in high-dimensional state spaces where they are still a method of choice in the financial industry. Recently, Tensor Processing Units (TPUs) have provided considerable speedups and decreased the cost of running Stochastic Gradient Descent (SGD) in Deep Learning. After highlighting computational similarities between training neural networks with SGD and simulating stochastic processes, we ask in the present paper whether TPUs are accurate, fast and simple enough to use for financial Monte Carlo. Through a theoretical reminder of the key properties of such methods and thorough empirical experiments we examine the fitness of TPUs for option pricing, hedging and risk metrics computation. In particular we demonstrate that, in spite of the use of mixed precision, TPUs still provide accurate estimators which are fast to compute when compared to GPUs. We also show that the Tensorflow programming model for TPUs is elegant, expressive and simplifies automated differentiation.

arXiv:1906.02818•Jun 6, 2019•3 citations

Large Scale Quantum Chemistry with Tensor Processing Units

Ryan D. Pederson, John Kozlowski, Ruyi Song +8 more

We demonstrate the use of Googles cloud-based Tensor Processing Units (TPUs) to accelerate and scale up conventional (cubic-scaling) density functional theory (DFT) calculations. Utilizing 512 TPU cores, we accomplish the largest such DFT computation to date, with 247848 orbitals, corresponding to a cluster of 10327 water molecules with 103270 electrons, all treated explicitly. Our work thus paves the way toward accessible and systematic use of conventional DFT, free of any system-specific constraints, at unprecedented scales.

arXiv:2202.01255•Journal of Chemical Theory and Computation•Dec 12, 2022•29 citations

Efficient 2D Tensor Network Simulation of Quantum Systems

Yuchen Pang, Tianyi Hao, Annika Dugad +2 more

Simulation of quantum systems is challenging due to the exponential size of the state space. Tensor networks provide a systematically improvable approximation for quantum states. 2D tensor networks such as Projected Entangled Pair States (PEPS) are well-suited for key classes of physical systems and quantum circuits. However, direct contraction of PEPS networks has exponential cost, while approximate algorithms require computations with large tensors. We propose new scalable algorithms and software abstractions for PEPS-based methods, accelerating the bottleneck operation of contraction and refactorization of a tensor subnetwork. We employ randomized SVD with an implicit matrix to reduce cost and memory footprint asymptotically. Further, we develop a distributed-memory PEPS library and study accuracy and efficiency of alternative algorithms for PEPS contraction and evolution on the Stampede2 supercomputer. We also simulate a popular near-term quantum algorithm, the Variational Quantum Eigensolver (VQE), and benchmark Imaginary Time Evolution (ITE), which compute ground states of Hamiltonians.

arXiv:2006.15234•International Conference for High Performance Computing, Networking, Storage and Analysis (SC)•Nov 1, 2020•17 citations
Tensor Processing Units (TPUs) were developed by Google exclusively to support large-scale machine learning tasks. TPUs can, however, also be used to accelerate and scale up other computationally demanding tasks. In this paper we repurpose TPUs for the challenging problem of simulating quantum spin systems. Consider a lattice model made of NNN spin-12\frac{1}{2}21​ quantum spins, or qubits, with a Hamiltonian H=∑ihiH = \sum_i h_iH=∑i​hi​ that is a sum of local terms hih_ihi​ and a wavefunction ∣Ψ⟩|Ψ\rangle∣Ψ⟩ consisting of 2N2^N2N complex amplitudes. We demonstrate the usage of TPUs for both (i) computing the ground state ∣Ψgs⟩|Ψ_{gs}\rangle∣Ψgs​⟩ of the Hamiltonian HHH, and (ii) simulating the time evolution ∣Ψ(t)⟩=e−itH∣Ψ(0)⟩|Ψ(t)\rangle=e^{-itH}|Ψ(0)\rangle∣Ψ(t)⟩=e−itH∣Ψ(0)⟩ generated by this Hamiltonian starting from some initial state ∣Ψ(0)⟩|Ψ(0)\rangle∣Ψ(0)⟩. The bottleneck of the above tasks is computing the product H∣Ψ⟩H |Ψ\rangleH∣Ψ⟩, which can be implemented with remarkable efficiency utilising the native capabilities of TPUs. With a TPU v3 pod, with 2048 cores, we simulate wavefunctions ∣Ψ⟩|Ψ\rangle∣Ψ⟩ of up to N=38N=38N=38 qubits. The dedicated matrix multiplication units (MXUs), the high bandwidth memory (HBM) on each core, and the fast inter-core interconnects (ICIs) together provide performance far beyond the capabilities of general purpose processors.
Tensor processing units (TPUs) are specialized hardware accelerators developed by Google to support large-scale machine-learning tasks but they can also be leveraged to accelerate and scale other linearalgebra-intensive computations.In this paper, we demonstrate the usage of TPUs for massively parallel classical simulations of quantum many-body dynamics on long time scales.We apply our methods to study the phenomenon of Floquet prethermalization, i.e., exponentially slow heating in quantum spin chains subject to high-frequency periodic driving.We simulate the dynamics of L = 34 qubits for over 10 5 Floquet periods, corresponding to circuits with 4 × 10 6 nearest-neighbor two-qubit gates.The circuits simulated have no additional symmetries and represent a pure-state evolution in the full 2 L -dimensional Hilbert space.This is achieved by distributing the computation over 128 TPU cores.On that size TPU cluster, we find speed-ups in wall-clock run time of 230 times and 15 times when compared to reference CPU and single-graphics-processing-unit (GPU) simulations, respectively, for shorter-time 30-qubit simulations that can be handled by all three platforms.We study the computational cost of the simulations, as a function of both the number of qubits and the number of TPU cores used, up to our maximum capacity of L = 40 qubits, which requires a "full pod" of 2048 TPU cores with tens of terabytes of memory in total.For these simulations, an eight-TPU-core machine is comparable to a single A100 GPU and thus the full TPU pod is comparable to a machine with hundreds of top-of-the-line GPUs.However, the TPU pod is more energy and cost efficient and readily accessible (via Google Cloud), unlike such large many-GPU configurations.We also study the accumulation of numerical error as a function of circuit depth in very deep circuits.Our work demonstrates that TPUs can offer significant advantages for state-of-the-art simulations of quantum many-body dynamics.
Abstract Researchers have put a lot of effort into reducing the gap between current quantum processing units (QPU) capabilities and their potential supremacy.One approach is to keep supplementary computations in the CPU, and use QPU only for the core of the problem. In this work, we address the complexity of quantum algorithm of arbitrary quantum state initialization, an important building block of quantum data analysis and machine learning.QPUs do not outperform classical machines with existing precise initialization algorithms.Hence, many studies propose an approximate but robust quantum state initialization.Cutting a quantum state into a product of (almost) independent partitions with the help of CPU reduces the number of two-qubit gates, and correspondingly minimizes the loss of state fidelity in the quantum part of the algorithm.To find the least entangled qubits, current methods compute the singular value decomposition (SVD) for each qubit separately with CPU.In this paper, we optimize CPU usage and memory resource bottlenecks.We consider Tucker tensor decomposition as an alternative to the CPU-based SVD in a single low-entangled qubit detection task without the loss of solution quality.Both proposed methods outperform the SVD in time and memory for systems of at least ten qubits. We achieve an order faster implementation and two orders less memory usage for a system of 15 qubits.
Monte Carlo methods are critical to many routines in quantitative finance such as derivatives pricing, hedging and risk metrics. Unfortunately, Monte Carlo methods are very computationally expensive when it comes to running simulations in high-dimensional state spaces where they are still a method of choice in the financial industry. Recently, Tensor Processing Units (TPUs) have provided considerable speedups and decreased the cost of running Stochastic Gradient Descent (SGD) in Deep Learning. After highlighting computational similarities between training neural networks with SGD and simulating stochastic processes, we ask in the present paper whether TPUs are accurate, fast and simple enough to use for financial Monte Carlo. Through a theoretical reminder of the key properties of such methods and thorough empirical experiments we examine the fitness of TPUs for option pricing, hedging and risk metrics computation. In particular we demonstrate that, in spite of the use of mixed precision, TPUs still provide accurate estimators which are fast to compute when compared to GPUs. We also show that the Tensorflow programming model for TPUs is elegant, expressive and simplifies automated differentiation.
Monte Carlo methods are critical to many routines in quantitative finance such as derivatives pricing, hedging and risk metrics. Unfortunately, Monte Carlo methods are very computationally expensive when it comes to running simulations in high-dimensional state spaces where they are still a method of choice in the financial industry. Recently, Tensor Processing Units (TPUs) have provided considerable speedups and decreased the cost of running Stochastic Gradient Descent (SGD) in Deep Learning. After highlighting computational similarities between training neural networks with SGD and simulating stochastic processes, we ask in the present paper whether TPUs are accurate, fast and simple enough to use for financial Monte Carlo. Through a theoretical reminder of the key properties of such methods and thorough empirical experiments we examine the fitness of TPUs for option pricing, hedging and risk metrics computation. In particular we demonstrate that, in spite of the use of mixed precision, TPUs still provide accurate estimators which are fast to compute when compared to GPUs. We also show that the Tensorflow programming model for TPUs is elegant, expressive and simplifies automated differentiation.
We demonstrate the use of Googles cloud-based Tensor Processing Units (TPUs) to accelerate and scale up conventional (cubic-scaling) density functional theory (DFT) calculations. Utilizing 512 TPU cores, we accomplish the largest such DFT computation to date, with 247848 orbitals, corresponding to a cluster of 10327 water molecules with 103270 electrons, all treated explicitly. Our work thus paves the way toward accessible and systematic use of conventional DFT, free of any system-specific constraints, at unprecedented scales.
Simulation of quantum systems is challenging due to the exponential size of the state space. Tensor networks provide a systematically improvable approximation for quantum states. 2D tensor networks such as Projected Entangled Pair States (PEPS) are well-suited for key classes of physical systems and quantum circuits. However, direct contraction of PEPS networks has exponential cost, while approximate algorithms require computations with large tensors. We propose new scalable algorithms and software abstractions for PEPS-based methods, accelerating the bottleneck operation of contraction and refactorization of a tensor subnetwork. We employ randomized SVD with an implicit matrix to reduce cost and memory footprint asymptotically. Further, we develop a distributed-memory PEPS library and study accuracy and efficiency of alternative algorithms for PEPS contraction and evolution on the Stampede2 supercomputer. We also simulate a popular near-term quantum algorithm, the Variational Quantum Eigensolver (VQE), and benchmark Imaginary Time Evolution (ITE), which compute ground states of Hamiltonians.