Lawrence Livermore National Laboratory

Creating a Single Comprehensive Algorithm for Quantum Computing

Delicate quantum computing systems are easily compromised by "noise." A new protocol from LLNL researchers could change that.

The impact of noise on quantum computing

While classical computing relies on 0s and 1s to process and store information, quantum computers use the principles of quantum physics to explode those binary limitations. With quantum computing, data can be stored in four states at the same time for powerful, flexible, and fast processing.

The current goal of quantum information science (QIS) is to solve quantum mechanical problems such as describing how two nucleons evolve as they interact. But moving from theoretical to applied quantum computing is complex, in part because formal quantum computing algorithms assume an idealized experimental environment, including noise-free hardware performance.

In practice, though, quantum simulations are subject to a variety of uncontrolled environmental interactions (aka, noise) due to factors like temperature fluctuations and cosmic rays. Compounding the problem, quantum computing algorithms rely on a series of gates—discrete preset logical operations—that each introduce the opportunity for noise.

Limiting and harnessing quantum noise

Our project proposes to disrupt QIS with an unconventional protocol designed to operate and even thrive with inherent or intentionally introduced quantum noise.

Our approach starts with creating a single comprehensive gate to capture the quantum dynamics of the simulations we run. Reducing the number of gates in the algorithm should reduce the noise introduced into the simulation.

We'll then use LLNL's testbed to demonstrate quantum simulations with real-time propagation techniques. We'll take things one step further by leveraging the dissipative mechanisms of the remaining noise to isolate the ground state of the simulated system.

Pushing quantum computing forward

Building on a successful demonstration of a two-body simulation, we'll develop additional gates that can work with present-day quantum hardware and that scale with the number of particles.

We'll then return to the LLNL testbed to validate those algorithms as potential solutions to the many-body problem and develop a co-design pathway for scaling to larger nuclei and increasing accuracy.

Together, these experiments will act as a springboard to revolutionize scientific computing by paving the way to exact quantum simulations of other complex microscopic systems and make quantum computing applicable to a broader class of problems.

Related Publications


  • Group Members Part of the project: Jonathan DuBois
  • Internal Collaborators: Sofia Quaglioni, Kyle Wendt, Kostas Kravvaris, Erich Ormand
  • External Collaborators: Piero Luci, Francesco Turro, Valentina Amitrano, Francesco Pederiva Affiliation: Physics Department, University of Trento, Via Sommarive 14, I-38123 Trento, Italy INFN-TIFPA Trento Institute of Fundamental Physics and Applications, Via Sommarive, 14, I-38123 Trento, Italy
  • Funding info: LLNL Disruptive Research LDRD, 3 years starting April 2019

A black-and-white graphic describing three types of nucleon-nucleon interaction.
A description of nucleon-nucleon interaction. From left to right: A single pion exchange, a spin-independent contract term, and a spin-dependent contract term.

A photograph of a metal transmon.
A 3-D transmon device used for quantum simulation.

A graphic depicting how four different nuclear states are represented via a quantum processing unit (QPU).
A map of nuclear states to quantum processing unit (QPU) states.

Two charts showing MHz over time (in nanoseconds) for the interaction of two neutrons.
The result of using an optimized algorithm with LLNL's 3-D transmon. The single gate enables the system to simulate spin Hamiltonian while reducing noise.

A chart showing decayed spin evolution over time.
A simulation of real time spin evolution using real device parameters. The colored circles represent the output probability over time. As the simulation progresses, evolution decays due to the decoherence of the transmon device.

 A chart comparing the power spectral density of a quantum simulation versus perfect spin evolution.
A comparison of the power spectral density of the quantum simulation (the dotted orange line) versus perfect spin evolution (the solid blue line).