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

IEEE Access

QTN-Hybrid: A DSP-Free Non-Linear Tensor-Network Classifier for Low-Latency FPGA Inference

Kyung-Won Cho, Sunghyun Go, Sungkwan Youm

Published Jan 1, 2026• 0 citations
Publisher

Abstract

Tensor-network (TN) classifiers based on matrix-product states (MPS) have recently emerged as a route to fully DSP-free inference on FPGAs, but to the best of our knowledge, every reported deployment has been strictly multilinear. We address this gap by introducing QTN-Hybrid, a quantum-inspired tensor-network (QTN) hybrid design in which a DSP-free non-linear MPS classifier inserts a LUT-fabric-mapped Hardtanh activation (implemented as two signed comparators and a three-way selection multiplexer) between bond contractions. The Hardtanh non-linearity bounds the hidden vector after every bond contraction and thereby confines the fixed-point dynamic range along the chain, so that accuracy is retained at low precision while hardware resources are reduced at the same time. The proposed model is trained on Fashion-MNIST at an operating point comparable in size to the published cascaded spaced-matrix-product-operator (CSMPO) reference design ( $\chi =4$ , thirty-two sites, $\approx 1{\,}$ k parameters), repeated over five random seeds, and evaluated after post-training quantization at precisions from 4–16 bits. At every precision QTN-Hybrid is more accurate than the linear MPS, by + 4.38–4.50 pp at 8–16 bits and by + 5.30 pp at 6 bits, and at 6 bits it remains above the 16-bit linear baseline. On the hardware side, the kernels are C-synthesised by high-level synthesis at a $250{\,}$ MHz target clock and the generated RTL is logic-synthesised to obtain resource and timing figures. The resulting designs use zero DSP blocks and require several times fewer LUTs than the linear 16-bit baseline, and both the 8-bit and the 6-bit configurations pass post-synthesis resource-capacity and timing checks on a mid-range FPGA and on a low-cost edge system-on-module, indicating at the post-synthesis level that deployment on small FPGAs is feasible, while routability and on-board operation remain to be verified.

DOI
10.1109/access.2026.3745871

Related papers

Towards Tensor Network Models for Low-Latency Jet Tagging on FPGAs

Alberto Coppi, Ema Puljak, Lorenzo Borella +6 more

We present a systematic study of Tensor Network (TN) models $\unicode{x2013}$ Matrix Product States (MPS) and Tree Tensor Networks (TTN) $\unicode{x2013}$ for real-time jet tagging in high-energy physics, with a focus on low-latency deployment on Field Programmable Gate Arrays (FPGAs). Motivated by the strict requirements of the HL-LHC Level-1 trigger system, we explore TNs as compact and interpretable alternatives to deep neural networks. Using low-level jet constituent features, our models achieve competitive performance compared to state-of-the-art deep learning classifiers. We investigate post-training quantization to enable hardware-efficient implementations without degrading classification performance or latency. The best-performing models are synthesized to estimate FPGA resource usage, latency, and memory occupancy, demonstrating sub-microsecond latency and supporting the feasibility of online deployment in real-time trigger systems. Overall, this study highlights the potential of TN-based models for fast and resource-efficient inference in low-latency environments.

cs.LGCross-listedarXiv:2601.10801•Jan 19, 2026

Tree Tensor Network implemented on FPGA as ultra-low latency binary classifiers.

Lorenzo Borella, Alberto Coppi, J. Pazzini +3 more

Tensor Networks (TNs) are a computational framework traditionally used to model quantum many-body systems. Recent research has demonstrated that TNs can also be effectively applied to Machine Learning (ML) tasks, producing results comparable to conventional supervised learning methods. In this work, we investigate the use of Tree Tensor Networks (TTNs) for high-frequency real-time applications by harnessing the low-latency capabilities of Field-Programmable Gate Arrays (FPGAs). We present the implementation of TTN classifiers on FPGA hardware, optimized for performing inference on classical ML benchmarking datasets. Various degrees of parallelization are explored to evaluate the trade-offs between resource utilization and algorithm latency. By deploying these TTNs on a hardware accelerator and utilizing an FPGA integrated into a server, we fully offload the TTN inference process, demonstrating the system’s viability for real-time ML applications.

International Conference on High Energy Physics (ICHEP)•Jan 27, 2025•0 citations

Ultra-low latency quantum-inspired machine learning predictors implemented on FPGA

Lorenzo Borella, Alberto Coppi, Andrea Triossi +4 more

Abstract Tensor networks (TNs) are a computational paradigm used for representing quantum many-body systems. Recent works have shown how TNs can also be applied to perform machine learning (ML) tasks, yielding comparable results to standard supervised learning techniques. In this work, we study the use of Tree TNs (TTNs) in high-frequency real-time applications by exploiting the low-latency hardware of the field-programmable gate array (FPGA) technology. We present different implementations of TTN classifiers, capable of performing inference on classical ML datasets as well as on complex physics data. A preparatory analysis of bond dimensions and weight quantization is realized in the training phase, together with entanglement entropy and correlation measurements, that help setting the choice of the TTN architecture. The generated TTNs are then deployed on a hardware accelerator; using an FPGA integrated into a server, the inference of the TTN is completely offloaded. Eventually, a classifier for high energy physics applications is implemented and executed fully pipelined with sub-microsecond latency.

arXiv:2409.16075•Machine Learning Science and Technology•Dec 10, 2025•15 citations

Quantized NN Workflow for Low-Latency Multi-Qubit State Discrimination on FPGA

Pradeep Kumar Gautam, Shantharam Kalipatnapu, H Shankaranarayanan +4 more

Measuring a qubit state is a fundamental but error-prone task in quantum computing. As the number of qubits increases, scalable readout demands multiplexed qubit measurement, which often suffers from crosstalk and other nonidealities. The conventional approach of qubit state detection often struggles with scalability and noise resilience. Neural network (NN)-based discriminators have emerged as a promising alternative, offering superior performance for inferring qubit states from multiplexed signals. However, their deployment on hardware platforms like Field programmable gate arrays (FPGA) is hindered by computation complexity, resource constraints, and latency. In this work, we present an end-to-end methodology for designing and deploying quantized NN-based multi-qubit state discriminators on FPGAs. We demonstrate that implementing a fully connected neural network accelerator for multi-qubit readout is advantageous, balancing computational complexity with low latency requirements without significant loss in accuracy. We significantly reduce the FPGA footprint and latency of the neural network by employing quantization-aware training (QAT) to quantize weights, activations, and inputs, and leveraging automated mapping flows for efficient, hardware-aware design-space exploration and optimization. The hardware accelerator performs frequency-multiplexed readout of five superconducting qubits in 26.7 ns, on a radio frequency system on chip (RFSoC) ZCU111 FPGA. These modules can be implemented and integrated into existing FPGA based quantum control and readout platforms, making them ready for experimental deployment.

International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies (HEART)•Jun 17, 2026•0 citations

QubiCML: ML-Powered Real-Time Quantum State Discrimination Enabling Mid-Circuit Measurements

Neel R Vora, Yilun Xu, Akel Hashim +11 more

This work presents QubiCML, a novel FPGA-based system engineered for real-time, in-situ discrimination of two- and three-level quantum states (quitrits) in superconducting quantum processors. Accurate and low-latency identification of quantum states is a critical enabler for scaling quantum computing, particularly as higher-dimensional systems such as qutrits are explored to increase information density, circuit depth, and gate expressivity. QubiCML integrates a lightweight, multi-layer neural network directly into FPGA hardware to classify the qutrit states$\vert g\rangle,\vert e\rangle$, and$\vert f\rangle$with high fidelity and minimal latency. Traditional state discrimination methods rely on offline post-processing, resulting in delays incompatible with the short coherence times of superconducting qubits. QubiCML overcomes this bottleneck by performing all inference operations within the FPGA fabric itself, achieving full quantum state discrimination within 66 ns (with softmax) or 40 ns (without softmax). The system supports fast reconfiguration via BRAM-stored parameters, enabling compatibility with varying quantum hardware and dynamic calibration environments. Additionally, we propose FPGA-friendly techniques for data normalization and fixed-point quantization to eliminate the overhead of floating-point and division operations—further optimizing the architecture for real-time deployment. We evaluate QubiCML on in-house superconducting hardware across multiple qubits and qutrits. The system achieves up to 97 % average accuracy for qutrits and up to 99 % for two-level systems, using readout windows as short as 650-750 ns. These results establish QubiCML as one of the first practical implementations of real-time qutrit state discrimination, marking an essential step toward mid-circuit measurements, dynamic feedback, and advanced error correction in higher-dimensional quantum computing.

IEEE International Conference on Quantum Computing and Engineering (QCE)•Aug 30, 2025•0 citations

Adaptive Tensor Network Sampling for Quantum Optimal Control

Zeki Zeybek, Rick Mukherjee, Peter Schmelcher

Quantum optimal control (QOC) provides a systematic framework for achieving high-fidelity operations in quantum systems and plays a central role in tasks such as gate synthesis, state transfer, and pulse design. Existing QOC methods broadly fall into two categories: gradient-based and gradient-free algorithms. The associated optimization landscape is often high-dimensional, non-convex, and populated by numerous local minima, making efficient gradient-free search strategies essential. To address this, we introduce a gradient-free matrix product state/tensor train (MPS/TT) sampling heuristic for discrete quantum optimal control. In our approach, the MPS defines a score function over the space of discrete control parameters, which in turn induces a sampling distribution over candidate control sequences. This distribution is iteratively refined through selection of better performing sequences and local tensor updates to bias the search toward high-performing sequences. We evaluate the method on a range of benchmark problems, including single-qubit state transfer, Bell-pair preparation, qutrit gate implementation, and open-system population transfer. Across these tasks, the method exhibits stable convergence behavior and competitive empirical performance relative to established gradient-free baselines. These results suggest that tensor network sampling offers a viable heuristic framework for discrete quantum control.

quant-pharXiv:2604.24467•Apr 28, 2026•0 citations
Tensor-network (TN) classifiers based on matrix-product states (MPS) have recently emerged as a route to fully DSP-free inference on FPGAs, but to the best of our knowledge, every reported deployment has been strictly multilinear. We address this gap by introducing QTN-Hybrid, a quantum-inspired tensor-network (QTN) hybrid design in which a DSP-free non-linear MPS classifier inserts a LUT-fabric-mapped Hardtanh activation (implemented as two signed comparators and a three-way selection multiplexer) between bond contractions. The Hardtanh non-linearity bounds the hidden vector after every bond contraction and thereby confines the fixed-point dynamic range along the chain, so that accuracy is retained at low precision while hardware resources are reduced at the same time. The proposed model is trained on Fashion-MNIST at an operating point comparable in size to the published cascaded spaced-matrix-product-operator (CSMPO) reference design ( χ=4\chi =4χ=4 , thirty-two sites, ≈1 \approx 1{\,}≈1 k parameters), repeated over five random seeds, and evaluated after post-training quantization at precisions from 4–16 bits. At every precision QTN-Hybrid is more accurate than the linear MPS, by + 4.38–4.50 pp at 8–16 bits and by + 5.30 pp at 6 bits, and at 6 bits it remains above the 16-bit linear baseline. On the hardware side, the kernels are C-synthesised by high-level synthesis at a 250 250{\,}250 MHz target clock and the generated RTL is logic-synthesised to obtain resource and timing figures. The resulting designs use zero DSP blocks and require several times fewer LUTs than the linear 16-bit baseline, and both the 8-bit and the 6-bit configurations pass post-synthesis resource-capacity and timing checks on a mid-range FPGA and on a low-cost edge system-on-module, indicating at the post-synthesis level that deployment on small FPGAs is feasible, while routability and on-board operation remain to be verified.
We present a systematic study of Tensor Network (TN) models $\unicode{x2013}$ Matrix Product States (MPS) and Tree Tensor Networks (TTN) $\unicode{x2013}$ for real-time jet tagging in high-energy physics, with a focus on low-latency deployment on Field Programmable Gate Arrays (FPGAs). Motivated by the strict requirements of the HL-LHC Level-1 trigger system, we explore TNs as compact and interpretable alternatives to deep neural networks. Using low-level jet constituent features, our models achieve competitive performance compared to state-of-the-art deep learning classifiers. We investigate post-training quantization to enable hardware-efficient implementations without degrading classification performance or latency. The best-performing models are synthesized to estimate FPGA resource usage, latency, and memory occupancy, demonstrating sub-microsecond latency and supporting the feasibility of online deployment in real-time trigger systems. Overall, this study highlights the potential of TN-based models for fast and resource-efficient inference in low-latency environments.
Tensor Networks (TNs) are a computational framework traditionally used to model quantum many-body systems. Recent research has demonstrated that TNs can also be effectively applied to Machine Learning (ML) tasks, producing results comparable to conventional supervised learning methods. In this work, we investigate the use of Tree Tensor Networks (TTNs) for high-frequency real-time applications by harnessing the low-latency capabilities of Field-Programmable Gate Arrays (FPGAs). We present the implementation of TTN classifiers on FPGA hardware, optimized for performing inference on classical ML benchmarking datasets. Various degrees of parallelization are explored to evaluate the trade-offs between resource utilization and algorithm latency. By deploying these TTNs on a hardware accelerator and utilizing an FPGA integrated into a server, we fully offload the TTN inference process, demonstrating the system’s viability for real-time ML applications.
Abstract Tensor networks (TNs) are a computational paradigm used for representing quantum many-body systems. Recent works have shown how TNs can also be applied to perform machine learning (ML) tasks, yielding comparable results to standard supervised learning techniques. In this work, we study the use of Tree TNs (TTNs) in high-frequency real-time applications by exploiting the low-latency hardware of the field-programmable gate array (FPGA) technology. We present different implementations of TTN classifiers, capable of performing inference on classical ML datasets as well as on complex physics data. A preparatory analysis of bond dimensions and weight quantization is realized in the training phase, together with entanglement entropy and correlation measurements, that help setting the choice of the TTN architecture. The generated TTNs are then deployed on a hardware accelerator; using an FPGA integrated into a server, the inference of the TTN is completely offloaded. Eventually, a classifier for high energy physics applications is implemented and executed fully pipelined with sub-microsecond latency.
Measuring a qubit state is a fundamental but error-prone task in quantum computing. As the number of qubits increases, scalable readout demands multiplexed qubit measurement, which often suffers from crosstalk and other nonidealities. The conventional approach of qubit state detection often struggles with scalability and noise resilience. Neural network (NN)-based discriminators have emerged as a promising alternative, offering superior performance for inferring qubit states from multiplexed signals. However, their deployment on hardware platforms like Field programmable gate arrays (FPGA) is hindered by computation complexity, resource constraints, and latency. In this work, we present an end-to-end methodology for designing and deploying quantized NN-based multi-qubit state discriminators on FPGAs. We demonstrate that implementing a fully connected neural network accelerator for multi-qubit readout is advantageous, balancing computational complexity with low latency requirements without significant loss in accuracy. We significantly reduce the FPGA footprint and latency of the neural network by employing quantization-aware training (QAT) to quantize weights, activations, and inputs, and leveraging automated mapping flows for efficient, hardware-aware design-space exploration and optimization. The hardware accelerator performs frequency-multiplexed readout of five superconducting qubits in 26.7 ns, on a radio frequency system on chip (RFSoC) ZCU111 FPGA. These modules can be implemented and integrated into existing FPGA based quantum control and readout platforms, making them ready for experimental deployment.
This work presents QubiCML, a novel FPGA-based system engineered for real-time, in-situ discrimination of two- and three-level quantum states (quitrits) in superconducting quantum processors. Accurate and low-latency identification of quantum states is a critical enabler for scaling quantum computing, particularly as higher-dimensional systems such as qutrits are explored to increase information density, circuit depth, and gate expressivity. QubiCML integrates a lightweight, multi-layer neural network directly into FPGA hardware to classify the qutrit states∣g⟩,∣e⟩\vert g\rangle,\vert e\rangle∣g⟩,∣e⟩, and∣f⟩\vert f\rangle∣f⟩with high fidelity and minimal latency. Traditional state discrimination methods rely on offline post-processing, resulting in delays incompatible with the short coherence times of superconducting qubits. QubiCML overcomes this bottleneck by performing all inference operations within the FPGA fabric itself, achieving full quantum state discrimination within 66 ns (with softmax) or 40 ns (without softmax). The system supports fast reconfiguration via BRAM-stored parameters, enabling compatibility with varying quantum hardware and dynamic calibration environments. Additionally, we propose FPGA-friendly techniques for data normalization and fixed-point quantization to eliminate the overhead of floating-point and division operations—further optimizing the architecture for real-time deployment. We evaluate QubiCML on in-house superconducting hardware across multiple qubits and qutrits. The system achieves up to 97 % average accuracy for qutrits and up to 99 % for two-level systems, using readout windows as short as 650-750 ns. These results establish QubiCML as one of the first practical implementations of real-time qutrit state discrimination, marking an essential step toward mid-circuit measurements, dynamic feedback, and advanced error correction in higher-dimensional quantum computing.
Quantum optimal control (QOC) provides a systematic framework for achieving high-fidelity operations in quantum systems and plays a central role in tasks such as gate synthesis, state transfer, and pulse design. Existing QOC methods broadly fall into two categories: gradient-based and gradient-free algorithms. The associated optimization landscape is often high-dimensional, non-convex, and populated by numerous local minima, making efficient gradient-free search strategies essential. To address this, we introduce a gradient-free matrix product state/tensor train (MPS/TT) sampling heuristic for discrete quantum optimal control. In our approach, the MPS defines a score function over the space of discrete control parameters, which in turn induces a sampling distribution over candidate control sequences. This distribution is iteratively refined through selection of better performing sequences and local tensor updates to bias the search toward high-performing sequences. We evaluate the method on a range of benchmark problems, including single-qubit state transfer, Bell-pair preparation, qutrit gate implementation, and open-system population transfer. Across these tasks, the method exhibits stable convergence behavior and competitive empirical performance relative to established gradient-free baselines. These results suggest that tensor network sampling offers a viable heuristic framework for discrete quantum control.