QxSQA: GPGPU-Accelerated Simulated Quantum Annealer within a Non-Linear Optimization and Boltzmann Sampling Framework

The QxBranch team presents QxSQA, our high-performance Simulated Quantum Annealer tuned for solving hard integer non-linear optimization problems and for deep learning using Boltzmann-based neural networks. Running on off-the-shelf consumer GPGPU hardware, QxSQA is integrated directly within the QxBranch Developer Platform, positioning users to rapidly develop, analyze, and deploy enterprise-grade applications for near-term quantum annealing and universal quantum computing devices.

9 April 2019 arXiv

Quanvolutional Neural Networks: Powering Image Recognition with Quantum Circuits

QxBranch Senior Data Scientist, Max Henderson et al. introduce a novel quantum algorithm with the potential of higher accuracy than traditional convolutional neural networks (CNNs). Their research shows that by adding a new quantum convolution (quanvolution) layer they can exceed the capabilities of normal CNNs.

1 March 2019 Journal of the Physical Society of Japan

Leveraging Quantum Annealing for Election Forecasting

The QxBranch team addresses the challenge of generating accurate election forecasts with current polling models. This paper investigates the application of quantum annealing to train Boltzmann machines using the 2016 US presidential election as a use case. 

25 June 2018 Adiabatic Quantum Computing Conference

Image Stacking by Quantum Annealing

QxBranch Systems Engineer, Tristan Cook, addresses challenges of image stacking using a quantum algorithm to build stacks based on correlated images. While there are some classical algorithms used to
select these subsets of images, it is a hard problem and most efficient methods are approximate by necessity. 

26 June 2017 Adiabatic Quantum Computing Conference

Quantum-Assisted Learning for Convolutional Deep Belief Networks

QxBranch Senior Data Scientist, Max Henderson, served as an advisor to Gabriel Bianconi and Mykel Kochenderfer of Stanford University’s Department of Computer Science, who investigate whether adiabatic quantum computers can improve deep learning algorithms and overcome existing limitations. They present a new quantum-assisted algorithm for convolutional deep belief networks with the goal of overcoming these limitations and enabling larger problems to be simulated.

14 September 2016 IEEE High Performance Extreme Computing Conference

Rapid Prototyping with Symbolic Computation: Fast Development of Quantum Annealing Solutions

In the current early stage of quantum computer hardware development, testing quantum algorithms on small scale hardware is critical. In this paper, members of the QxBranch team present a software library of symbolic computation functions that makes this prototyping possible. 

29 June 2015 Adiabatic Quantum Computing Conference

A Novel Embedding Technique for Optimization Problems of Fully-Connected Integer Variables

QxBranch software engineers along with Kenneth Zick of the University of Southern California Information Sciences Institute, presented on their work that illustrates the potential for broadening the variety of problems that can be mapped to adiabatic quantum computers. This poster showcases their solution and findings.

Subscribe to our newsletter