Reducing Signal-to-Noise and Distortion with Radio Frequency Photonics

LLNL researchers are advancing quantum computing by refining the classical computing components used in classical-quantum interfaces.

The meeting of quantum and classical computing

Present-day quantum computing (QC) is accomplished via a combination of classical and quantum hardware. For example, although QC data is stored on qubits (quantum bits), the systems that operate and read the data rely on classical computing platforms.

This interface needs to provide high-fidelity wideband signals to control and measure quantum devices, but it can introduce noise, heat, and restrictions on the overall size and scale of the device.

To balance these undesirable effects, researchers often restrict device operations, but that comes at a cost. For example, setting an upper bound on gate speed reduces noise but can negatively impact processing time.

Improving the classical-quantum interface with photonics

To improve the scalability and performance of superconducting quantum computing systems, we are building a new classical-quantum interface based on RF photonics.

Conventional digital-to-analog converters rely on electronic upconversion to tune the amplitude and phase of generated radio frequency tones. Although the tuning is necessary, conventional converters are limited by mixer distortion that can compromise quantum experiments.

In contrast, a photonic digital-to-analog converter (PDAC) optically generates RF tones from a train of laser pulses that use electro-optical sampling. The ultrashort duration of these pulses suppresses RF jitter noise, providing excellent phase stability for the resultant RF signals.

Put another way, by sampling the lowest-noise portion of the electronic modulation, the PDAC is able to greatly improve signal-to-noise and distortion (SINAD). In addition to noise benefits, PDAC technology enables faster gates without sacrificing coherence.

Fine-tuning quantum computing hardware

Along with implementing PDAC converters, our project will reduce the heat load imparted by the control lines and the thermal background present control signal by replacing RF cabling with optical fiber.

By refining the classical-quantum interface, our project removes artificially imposed constraints on quantum processing and expands the possibilities of quantum computing.

Related Publications

  • Wideband, high fidelity RF signal generation using photonics | LLNL External Review Committee Poster Session, 2020 J. Chan, A. Gowda, P.T.S. DeVore, B.W. Buckley, J.L. DuBois, and J. Chou
  • High-fidelity, scalable quantum-classical control interface using photonics | American Physical Society March Meeting, Denver, Colorado, 2020 J. Chan, A. Gowda, P.T.S. DeVore, B.W. Buckley, J.L. DuBois, and J. Chou

Patents

Radio frequency passband signal generation using photonics | Non-Provisional Patent Application, 2020 A. Gowda, J.C.K. Chan, P.T.S. DeVore, D.S. Perlmutter, J.T. Chou

People

  • Group members that are part of the project: Kristi Beck, Kevin Chaves, Jonathan DuBois
  • Internal collaborators: Apurva Gowda, Jacky Chan, Peter DeVore, Jason Chou, Susant Patra, Gianpaulo Carosi
  • Funding info: LDRD ER, FY21-FY23
A graph showing the measured and simulated fidelity and bandwidth of an llnl team's pdac (photonic digital-to-analog converter) versus commercially available electronic signal generators.
A comparison of LLNL's innovative PDAC (photonic digital-to-analytic converter) versus multiple commercially available electronic signal generators.
A schematic showing a high-fidelity quantum drive signal generator. the graphic shows that the output of an electronic digital-to-analog converter array will be filtered through a photonic integrated circuit, undergo radio frequency modulation, then move through an optimized optical fiber array and a photodetector array before arriving at a multi-qubit quantum system.
Our new classical-quantum interface based on RF photonics reduces noise and distortion when compared with conventional digital-to-analog converters in the few GHz range relevant for superconducting qubits by filtering electronic digital-to-analog converters through an integrated photonic circuit.