Scaling Photonic Quantum Computers

Scaling photonic quantum computers

ZACHARY VERNON University of Pavia Colloquium 29 April 2021

1. ABOUT US 4. APPLICATIONS: NISQ AND FTQC

2. PHOTONICS FOR QUANTUM COMPUTING 5. NANOPHOTONIC DEVICES FOR SQUEEZING

CEO HARDWARE SOFTWARE ARCHITECTURE PRODUCT

Christian Weedbrook, PhD Zachary Vernon, PhD Nathan Killoran, PhD Ish Dhand, PhD Rafal Janik, MSc

Founded in 2016 | HQ in Toronto | 70+ people | 40+ PhDs | 15+ nationalities

Advisors:

Founded in 2016 Funding: $45M Ofﬁce & Hardware Lab in Toronto

Requirements What makes a qubit? 1. A scalable physical system with well DiVincenzo criteria, 2000: characterized qubit 2. The ability to initialize the state of the qubits to a simple ﬁducial state 3. Long relevant decoherence times 4. A "universal" set of quantum gates 5. A qubit-speciﬁc measurement capability Fortschr. Phys.4892000) 9-11, 771-783 Desiderata

1. The ability to interconvert stationary and flying qubits 2. The ability to faithfully transmit flying qubits between specified locations

Requirements

1. A scalable physical system with well Trapped ions Superconducting electronics characterized qubit 2. The ability to initialize the state of the qubits to a simple ﬁducial state 3. Long relevant decoherence times 4. A "universal" set of quantum gates 5. A qubit-speciﬁc measurement capability

Source: Physics World Source: Phys.org Desiderata

1. The ability to interconvert At the time (c. 2000), matter-based qubits were stationary and flying qubits generally assumed to be the only option 2. The ability to faithfully transmit flying qubits between specified Blocker: Lack of single-photon level nonlinearity locations

“KLM” protocol for photonic quantum computing, 2001:

Nature volume 409, pages 46–52(2001)

Showed that photonic QC was, in principle, possible without exotic photon-photon interactions.

Tradeoff: Lack of deterministic operation, leading to huge component counts.

Photonics Superconducting Trapped ions

Operating Mostly room temperature, no Cryostat @ 10mK Vacuum chamber, laser vacuum cooling to nK environment

Chip integration Compatible with standard Specialized Cannot be fully chip CMOS fabs research/university fabs integrated

Prospects for scaling Modularity via telecom No suitable interconnect yet Optical interconnect using ﬁber-optic interconnects demonstrated UV/visible photonic elements

Maturity of industry Beneﬁts from optical telecom Small market for Very little market for industry components superconducting electronics components outside for components in other applications scientiﬁc research

Bandwidth/clock Ultrahigh bandwidth: GHz kHz reset rates, gates can be Slow reset times and gate possible reasonably fast speeds speeds

Photonics Superconducting Trapped ions

Platform simplicity Difﬁcult to do everything in Qubits, gates, measurements Complicated laser control one material platform, need done in same layer, control (many UV/visible beams heterogeneous and readout done with RF required) integration/multi-chip pulses architecture

Component counts Large due to Moderate, but wiring inside Low, also due to high non-deterministic elements cryostat is a serious challenge ﬁdelities allowing possibly (cryoCMOS required?) lower overheads for fault tolerance

FASTER UNIQUE LOWER SCALABLE NETWORK DEVELOPMENT + ENCODING + COST + & MODULAR + COMPATIBLE

Much faster Different NISQ applications Fiber interconnects Can borrow many tools from Natural interface with prototyping cycles due to room and more ways to implement distribute qubits across optical telecom industry future quantum internet temperature QPUs error correction many chips

9 XANADU’S ROADMAP TO UNIVERSAL QUANTUM COMPUTING

NISQ GATE-BASED SYSTEMS

GAUSSIAN Xanadu’s BOSON GKP QUBITS approach SAMPLING UNIVERSAL FAULT-TOLERANT QUANTUM COMPUTER

Other approaches VQE

QAOA

Time

The formula

Squeezed states (single temporal mode)

+ Linear optics = (programmable, phase stable)

+ Photon counting (photon number resolving)

Source: “Applications of near-term photonic quantum computers: Software and algorithms”, T.R. Bromley et al., Quantum Science and Technology (2020)

Source: Science 18 Dec 2020: Vol. 370, Issue 6523, pp. 1460-1463

Source: “Applications of near-term photonic quantum computers: Software and algorithms”, T.R. Bromley et al., Quantum Science and Technology (2020)

Input: Compiled program loaded to chip via DACs

Pulsed pump laser source

Output: Programmable 8-mode entangled Squeezed Programmable interferometer light sources (Information encoding and quantum gate sequence) Gaussian state

Equivalent quantum circuit

Input / Output Master Controller (Sends and receives classical information to the various components)

User Interface & Remote Access (Xanadu’s Strawberry Fields software) Laser pump Chip control system (optical source) (DACs to encode information)

QPU Data acquisition system Xanadu’s 1st generation chip

Single-squeezer statistics

Two-squeezer interference

DENSE MAXIMUM POINT SUBGRAPH CLIQUE PROCESS

QUANTUM GRAPH VIBRONIC ENHANCED SIMILARITY SPECTRA FEATURE

“Applications of near-term photonic quantum TUTORIALS AVAILABLE AT STRAWBERRYFIELDS.AI computers: Software and algorithms”, T.R. Bromley et al., Quantum Science and Technology (2020)

● Randomized benchmark for device performance

● Compare experimental and theoretical distributions

● Good quantitative and qualitative agreement between experiment and theory

19 SIMULATION OF VIBRONIC SPECTRA

● Chemistry application unique to squeezing-based bosonic architecture

● Encode molecular structure directly in chip parameters

20 GRAPH SIMILARITY

● Graph-structured data encoded into X8 chip

● Readout statistics classify diﬀerent graphs

● Clustering shows X8 can distinguish diﬀerent graph structures

21 GBS FOR FAULT TOLERANT QUANTUM COMPUTING

state factory multiplexer ● Architecture relies on GKP computational module qubits, enabling two layers of error correction p-squeezed state ● X8 chips are the building blocks for synthesizing GKP qubits ● Squeezed states help deal with non-deterministic GKP sources

GKP state QPU

Blueprint for a Photonic Fault-Tolerant Quantum Computer J. Eli Bourassa et al. Quantum 5, 392 (2021)

22 NANOPHOTONIC SOURCES OF SQUEEZED LIGHT

Wish list for integrated squeezing platform & device

Low conﬁnement PPLN SiN OPOs above threshold SiN microring below threshold

Nanophotonic system X ✓ ✓ Quadrature squeezing ✓ X ✓ Photon counting compatible ✓ X ✓ Single temporal mode ? X ✓ No excess noise ✓ X ✓

Lenzini et al., Sci. Adv. 4, eaat9331 (2018)

Dutt et al., Phys. Rev. Vaidya et al., Sci. Adv. eaba9186 (2020) Applied 3, 044005 (2015) Zhang et al., Nat. Comm. (2021)

Zhang et al., “Squeezed Light from a Nanophotonic Molecule”Nature Communications (2021)

Record highest squeezing from a nanophotonic device!

Similar setup to before, HNLF added to Similar setup to before, HNLF added to generate a tighter phase lock generate a tighter phase lock

Zhang et al., “Squeezed Light from a Nanophotonic Molecule”Nature Communications (2021)

Full-time positions available - xanadu.ai/careers

[email protected]