Overview

State-of-the-art demonstrations – which have only achieved fidelities of 97% in free-space and 58% with cavities – require significant technical advances. Nu Quantum targets a very high system-level performance, with fidelity >99% and rate >4.5kHz  through the photonic network. Performance starts at the QPI: an open microcavity that allows to collect photons efficiently without deteriorating the qubit quality.

At Nu Quantum, we have made steady progress towards making this system-level performance a reality. Our QPI technology is based on ultra-high finesse optical microcavities. There are three pieces of core technology inside the QPI:

• Microcavity mirrors
  
• Active locking of the cavity length
  
• Spin-Photon Entanglement Protocol
  

We have now proven our technology platform by achieving the specifications on the three items above that will support Gen1-3 of our QPI products, to Fault Tolerant networking targets. We can now make the best microcavities in the world, we have the tightest locking and the world’s most powerful and robust entanglement protocol.

This core technology will be implemented into different generations of our QPI products. We have built Gen1 QPI, and it has been shipped and tested by a qubit partner (without qubits yet, optically) - it meets all specifications. Series A will allow us to test Gen1 QPI with a qubit.

Series A will also allow Nu Quantum to marry our unique cavity-fabrication workflows with wafer-scale ion-trap manufacturing: Gen2 QPI - going beyond prototype manufacturing and demonstrate that our high-performance QPI can be manufactured on an industrial scale.

Technology Platform

Open Micro-cavities

Nu Quantum’s microcavities are unique— our microcavity fabrication workflows yields perfectly spherical mirror, with high smoothness and with ~100μm radius-of-curvature.

The spherical shape with small radii is the optimal shape for quantum applications: we can create strong light-matter coupling without bringing the mirrors close to the qubits (which would perturb their quantum properties!).

There are many other approaches out-there: Open micro-cavities with parabolic mirrors, whispering gallery modes, photonic crystals… None allow to collect photons as speedily and efficiently as us whilst keeping the mirrors ~100μm away from the precious qubits!

Micro-cavity Mirrors

The fabrication of our cavity-mirrors is a trade-secret, but you can judge us on numbers!

• our mirrors can be manufactured on any flat substrate—wafers or conical pins.
• we position our mirrors with sub-micron accuracy.
• our mirrors are spherical in shape and the radii-of-curvature are controlled within a few percent accuracy.
• our surface quality is pristine, with <15ppm loss per coated-substrate, corresponding to cavity finesses, F>200,000.

Cavity Locking

A cavity is only good if it is resonant with the photons that our qubits emit! Nu Quantum has already demonstrated in past-projects that we could control the cavity resonance with remarkably high accuracy. For a neutral-atom QPI, the 2mm length of our cavity was locked down to ~80pm—on a human scale, this corresponds to the width of a human hair compared to the Empire-State Building.

• Our cavity-locking capabilities have been developed and productised in-house (see picture), and independently validated by two external partners.
• We have fully specified and patented a scalable way to lock a network of cavities-paving the way to the Entanglement Fabric.

Network Entanglement Protocol

Nu Quantum targets commercially relevant rates and our targets are informed by a bottom-up system-modelling of our hardware capabilities, such that they represent what our hardware can or will be able to do. We benchmark commercial relevance via our QEC error-model, combined with a commonly accepted metric: QEC cycles must be 1 ms or less to access utility-scale quantum computing. Our QEC model requires ~6 checks within a QEC cycle, which places a commercially relevant entanglement rate at 6 kHz. Our QEC model also places a bound on a commercially relevant Bell-pair remote entanglement fidelity at 99.5%. This error-model assumes that all other gates (which will be local to each QPU) operate at 99.99% fidelity—a regime that qubit companies are targeting and close to demonstrating with small QPU sizes (~100 qubits) . It is worth noting that these numbers will only continue to improve as QEC and hardware innovations improve.

The table below summarises Nu Quantum’s targets (for a trapped ion QPI) and how they compare relative to the current State-of-the-Art (SoA), demonstrated by academic groups at the University of Oxford (UoO) and University of Innsbruck (UoI).

{{entanglement-protocol="/rich-text-content/tables/entanglement-protocol"}}

From this table, it is very apparent that our targets go well-beyond the state of the art. The key blockers are:

• (Free-space) Can’t reach entanglement rates to support Distributed Quantum Computing.
• (Free-space & Cavity) Sub-optimal spin-photon entanglement schemes, with intrinsically low entanglement fidelity.
• (Cavity) Inability to perform advanced quantum networking experiments with micro-cavities due to system-engineering complexity.

One aspect to solving this problem lies at the QPI to get the best ion-photon interface, the other lies in quantum engineering, and improving the performance of Entanglement protocols. Lots of ‘little’ things went into the design of our entanglement protocol to ensure it would support high fidelities:

• Minimise photon Recoil: Use NIR-transition & work close to motional ground-state: Nu Quantum will use a Sr+ transition at 1033 nm.
• Use time-bin entanglement: Nu Quantum will use this encoding, allowing high-fidelity photon propagation in a fibre-network.
• Leverage a high-fidelity v-STIRAP configuration: Nu Quantum’s entanglement scheme is high-fidelity within a practical range of cavity-QED parameters (see figures below)

‍

Gen-1 QPI Prototype

Mechanical Assembly

Mechanical assembly has been the pitfall of many academic groups. At Nu Quantum, we have a dedicated mechanical engineering team, who have successfully developed integration strategies to meet the ~5µm accuracy required by our micro-cavities assemblies. These passively-aligned cavity assemblies are a world-first, and feature in-built mechanical stability.

Successful integration also requires our cavity assemblies to work in Ultra-High-Vacuum (UHV). This means that our assemblies have to survive the ‘bake-out’ process, and should only contain UHV-compatible materials.

Beyond these general integration principles, each cavity assembly must fulfil requirements that are qubit-specific. Nu-Quantum has developed a neutral atom QPI, where optical access on three axes and with six-beams was required. We have also developed in partnership with the University of Oxford a trapped-ion QPI, where the cavity mirrors needed to fit within metal electrodes.

To mature our technology further, we have decided to focus our Series A hardware developments on trapped-ion qubits. They have the most pressing networking needs, and we thus foresee huge traction once we stand up our quantum-network demonstrator. That said, our micro-fabrication and manufacturing skills and hugely transferrable to other qubit modalities.