Abstract:Accurately predicting a quantum computer's capability -- which circuits it can run and how well it can run them -- is a foundational goal of quantum characterization and benchmarking. As modern quantum computers become increasingly hard to simulate, we must develop accurate and scalable predictive capability models to help researchers and stakeholders decide which quantum computers to build and use. In this work, we propose a hardware-agnostic method to efficiently construct scalable predictive models of a quantum computer's capability for almost any class of circuits, and demonstrate our method using convolutional neural networks (CNNs). Our CNN-based approach works by efficiently representing a circuit as a three-dimensional tensor and then using a CNN to predict its success rate. Our CNN capability models obtain approximately a $1\%$ average absolute prediction error when modeling processors experiencing both Markovian and non-Markovian stochastic Pauli errors. We also apply our CNNs to model the capabilities of cloud-access quantum computing systems, obtaining moderate prediction accuracy (average absolute error around $2-5\%$), and we highlight the challenges to building better neural network capability models.

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve The paper aims to address the issue of predicting the capabilities of quantum computers. Specifically, the goal of the paper is to develop a method that can accurately predict the performance of quantum computers when executing specific quantum circuits. As modern quantum computers become increasingly difficult to simulate, researchers need to develop accurate and scalable capability prediction models to help researchers and stakeholders decide on the type of quantum computer to build and use. ### Main Problem Description 1. **Predicting the Capabilities of Quantum Computers**: - Accurately predict which circuits a quantum computer can run and the effectiveness of running these circuits. - These predictions are crucial for selecting the appropriate quantum computer to solve specific problems. 2. **Limitations of Existing Methods**: - Existing benchmarking methods (such as algorithm benchmarks and random circuit benchmarks) can only measure the performance of quantum processors on a small number of circuits and cannot predict their performance on untested circuits. - A new method is needed to efficiently construct scalable capability prediction models. 3. **Proposed Method**: - The paper proposes a hardware-agnostic method that uses Convolutional Neural Networks (CNN) to efficiently build prediction models for the capabilities of quantum computers. - This method represents quantum circuits as 3D tensors and then uses CNN to predict the success rate of the circuits. 4. **Application Scenarios**: - This method is applicable not only to simulated quantum processors but also to cloud-accessed quantum computing systems. - Experimental results show that this method achieves an average absolute prediction error of about 1% when dealing with circuits containing both Markovian and non-Markovian random Pauli errors. ### Summary The main goal of the paper is to develop a method that can accurately predict the capabilities of quantum computers to help researchers and stakeholders choose the appropriate quantum computer. By using Convolutional Neural Networks, the paper proposes a hardware-agnostic, efficient, and scalable method that can make accurate predictions across various types of quantum circuits. This method performs well not only in simulated environments but also in actual cloud-accessed quantum computing systems.

Learning a quantum computer's capability

Learning a quantum computer's capability using convolutional neural networks

What is my quantum computer good for? Quantum capability learning with physics-aware neural networks

Predictive Modelling of Quantum Process with Neural Networks

Predictive Models from Quantum Computer Benchmarks

Quantum Algorithms for Feedforward Neural Networks

A Quantum Leaky Integrate-and-Fire Spiking Neuron and Network

Neural Predictor Based Quantum Architecture Search

Benchmarking Quantum Convolutional Neural Networks for Classification and Data Compression Tasks

Quantum-inspired activation functions in the convolutional neural network

High Dimensional Quantum Machine Learning With Small Quantum Computers

Quantum-inspired activation functions and quantum Chebyshev-polynomial network

Predicting properties of quantum systems by regression on a quantum computer

Quantum Fourier Convolutional Network

Challenges and opportunities in the supervised learning of quantum circuit outputs

Quantum Convolutional Neural Networks are (Effectively) Classically Simulable

Quantum Neural Network for Quantum Neural Computing

Case study on quantum convolutional neural network scalability

QFCNN: Quantum Fourier Convolutional Neural Network

What can we learn from quantum convolutional neural networks?