By Dr David Shaw, GQI Chief Analyst
This week Photonic Inc, a Vancouver-based quantum startup, has announced details of the new quantum architecture it has been developing in stealth for the last seven years. It also announced that it has closed its latest $100M funding round, most strikingly with Microsoft Corporation as a prominent new investor. Photonic Inc and Microsoft have also announced a strategic collaboration aimed both at bringing this new technology to Microsoft Quantum Azure Elements, and also for its wider exploitation as a quantum networking technology. Taken together, GQI views these as the most significant developments so far this year in the quantum sector.
For a brief overview of Microsoft and Photonic Inc’s recent announcements [1]–[3] please read GQI’s summary.
In a world where popular articles report even rather routine quantum progress as major breakthroughs it can be hard to stand out from the crowd.
This is something different
To understand the disruptive potential of the new architecture we first have to recognize the silos into which the quantum world is currently split:
- Matter qubit enthusiasts point to their long qubit lifetimes;
… Photonic enthusiasts point to their potential for interconnects. - Trapped ion/atom specialists work at the short wavelengths those qubits demand;
… Networking specialists work at the longer wavelengths optic fibers support. - Silicon spin experts point to the great potential 28Si has to host long-lived spin qubits;
…Silicon photonics experts point to native quantum photonic support. - Quantum computing specialists highlight entanglement as a resource for computation;
… Quantum comms specialists focus on entanglement for cryptography. - Quantum sensing experts play on their own; no one has worked out how to link them up.
It can be argued that in practical terms all these different groups have been held apart by a missing technical component: an efficient spin-photon interface realizable in a readily manufacturable package and operating in a network friendly telecom band.
Founded in 2016, Photonic Inc went looking for the solution to this problem. Drawing on academic work dating back as far as 1981 and with a focus on research and development until 2020 [4]–[9]. Since 2021 it has been building its new combined quantum computing and networking architecture around the T center, a manufactured color center defect in silicon [10]–[12]. This brings a series of advantages:
- 28Si promises long qubit lifetimes, even at relatively modest temperatures (1-2 K) [7]
- T centers operate at c.1326 nm, compatible with the telecoms O-band [8], [9]
- SOI brings access to advanced nanofab technologies and components
- Waveguides and optical cavities to enhance the spin-photon interface [10], [11]
- Thousands of T centers can be fabricated in compact arrays [12]
A SOI-native spin-photon interface would already be a big deal. However, the new Photonic Inc architectural philosophy promises to bring more [13].
For a conventional architecture we would now ask, what mechanism can I use to form 2Q gates between locally adjacent computational qubits? How can I couple these computational qubits to communication qubits? The new architecture side steps these problems. None of the T center qubits are locally connected. All-to-all connectivity is provided by a flexibly switched optical network, with 2Q gates implemented via Bell pair measurement (the Barrett-Kok scheme [14]).
The proposed optical switch can be placed within the cryostat or at room temperature linking multiple cryostats. This opens up the possibility of an architecture that has:
- High modular scalability using native telecoms compatible links
…this contrasts with the challenge others face of generating a high-rate interface without the benefit of a waveguide-native qubits and/or the additional challenge of wavelength transduction. - Natural all-to-all connectivity; potentially allowing high-rate Q LDPC codes to be used for error correction.
… this contrasts with the high overheads faced by those proposing to use the 2D Surface Code (or other planar code variants)
If the key enabling components can indeed be realized with sufficient underlying performance, this architecture promises a conceptually straightforward route to what Photonic Inc calls Scalable Fault-Tolerant Quantum Computing (SFTQ) [13].
Challenges ahead
It’s one thing to have a blueprint for a great quantum architecture. It’s another thing to have the components to build it. How far along is Photonic Inc in realizing this journey?
Beyond the academic papers it has published on T centers, Photonic Inc is not currently releasing further details publicly. It does not currently offer any public access to its prototype devices, nor has it publicly promised a date for such access.
However, founder Dr Stephanie Simmons says, “we believe that—within five years, significantly sooner than the widely accepted timeframe—we will be the first quantum computing company to offer a scalable, distributed, and fault-tolerant solution.”
As part of the current investment round, GQI conducted technical due diligence and has looked in more detail at confidential Photonic Inc progress.
An in-depth, confidential 48-page technical due diligence report is available from GQI with prior permission from Photonic Inc management.
Assessed against our pre-existing quantum hardware stack framework we find Simmons’ statement to be ambitious but based on a plausible roadmap. The presence of knowledgeable quantum investors supporting the round is a further indication of the progress already made.
GQI identifies a series of key challenges that Photonic Inc must demonstrate it can address:
- Achieving a sufficiently high rate of entanglement generation between T center qubits; in this architecture fidelity and rate are ultimately a trade-off.
(the target rate is confidential) - Achieving sufficiently high component yields across its T center qubit arrays.
(the target module size is confidential) - Delivering a dynamic optical switching system at scale, while managing the need for its calibration and recalibration.
(The target switching rate is confidential) - Identifying a specific Q LDPC architecture with an associated universal fault tolerant gate set.
(the specific codes the team is currently studying are confidential) - Securing a source of silicon-28 for at-scale manufacture.
(the planned source is confidential)
If these elements can indeed be delivered, the new architecture has high modular scalability. GQI sees no obvious firewall between the realization of SFTQ and the creation of a cryptographically relevant quantum computer. We believe this architecture should cause those planning their migration to quantum safe cryptography to carefully reassess their ‘reasonable worst case’ Q-day.
The other clear opportunity for initial leverage of Photonic Inc’s SFTQ architecture is in quantum simulation applications (e.g. material science and quantum chemistry), or other problems where a strong quantum speedup is expected. It’s important to point out that Photonic Inc don’t make claims that their architecture will allow problems with only quadratic speedups to be addressed. This is not a turbo FTQC architecture.
The pull of entanglement networking
Entanglement is seen by many as the ultimate quantum resource, unlocking unique possibilities in quantum computing, quantum cryptography and quantum sensing. Today these pillars of quantum technology can seem operationally distinct. But it’s important to realize that this is because we currently lack the high-performance entanglement networking technology required for their unification.
Against this background the importance of the Microsoft – Photonic Inc collaboration is clear. It combines the capabilities of a cloud computing and software platforms powerhouse with a provider of potentially key enabling hardware.
Whereas T centers may be used to power QPU devices, they provide an even earlier path for the development of increasing sophisticated networking components:
- Memory assisted MDI-QKD – with the potential to offer Many-to-One networks.
- Telecom wavelength quantum repeaters – enabling long-distance quantum comms.
- Error corrected quantum repeaters – effectively small fault tolerant quantum devices.
This quantum internet technology promises a future where applications are blurred between networked quantum computing and advanced quantum cryptographic protocols such as blind computing; offering advanced functions from secure time synchronization to distributed coherent quantum sensing.
GQI feels it’s important to realize that quantum frameworks, platforms, and algorithms developed for conventional monolithic conceptions of FTQC may not find themselves optimized for this more network-centric, distribution friendly opportunity. Quantum software experts are going to have to be mobilized to get the most out of this new environment.
Microsoft’s recent blog post on quantum networking provides a framework for this journey [15]. In the short term it reflects the skepticism of the US NSA on first generation quantum security products (such as current generation QKD technology). In the medium term GQI is waiting to see how Microsoft’s current commitment to PQC evolves in line with the new quantum cryptographic opportunities it is investing in.
T centers vs topological qubits
Microsoft has its own established quantum computing hardware roadmap. It’s natural to ask how this is affected by its new investment and collaboration with Photonic Inc.
For GQI’s review see A Deeper Dive Into Microsoft’s Topological Quantum Computer Roadmap [16]
GQI views the two architectures as entirely complementary:
- Even the best conceived QC roadmaps remain high-risk ventures; it is perfectly prudent to be pursuing multiple options.
- The topological qubits roadmap has a much longer time horizon (GQI estimates perhaps 10-15 years vs 5 years). If it does manage to deliver it would provide qubits with higher raw fidelity and potentially higher gate speeds.
- Topological qubits (or any 3rd generation qubit technology) might still benefit from network components provided by T centers.
T centers vs the rest
What pressure will these new announcements put on others in the quantum hardware community?
Of course, both foreseen and unforeseen challenges may yet derail the delivery of Photonics’s SFTQ roadmap, which in the absence of published device stats remains at the PoC stage..
Even if Photonic Inc’s vision is realized, others will point to the potential advantages of their own architectures if they too can be delivered to similar scale:
- Superconducting circuit platforms will point to their higher logical cycle times they target
- Trapped Ions will point to potentially higher raw fidelities
- Photonic qubits will point to their own network friendly properties; and high cycle times
- Neutral atoms will point to how far they can push without interconnects
- Conventional silicon spin approaches will point to the dense qubit arrays they promise
- NV Diamonds will point to their potential for room temperature operation.
The true new pressure may be one of timescale. Investors will be less patient of longer roadmaps unless they have strong merits beyond early FTQC. Investors may see a shorter timeline to the payback offered by large FTQC applications; however, they will also see a reduced window for those targeting NISQ quantum advantage to demonstrate their worth.
The race is still a marathon, but the pace has started to ramp up.
References
[1] ‘Photonic Accelerating Quantum Computing’s Transformational Benefits with New Architecture’, Photonic. Accessed: Nov. 09, 2023. [Online]. Available: https://photonic.com/news/accelerating-quantum-computing-benefits-with-new-architecture/
[2] ‘Photonic Raises $100 Million USD for Quantum Technology from BCI, Microsoft, and Other Investors’, Photonic. Accessed: Nov. 09, 2023. [Online]. Available: https://photonic.com/news/photonic-raises-100m-for-quantum-technology/
[3] ‘Photonic Collaborating with Microsoft to Power Global Quantum Ecosystem’, Photonic. Accessed: Nov. 09, 2023. [Online]. Available: https://photonic.com/news/photonic-collaborating-with-microsoft-to-power-global-quantum-ecosystem/
[4] G. Davies, ‘The optical properties of luminescence centres in silicon’, Phys. Rep., vol. 176, no. 3, pp. 83–188, May 1989, doi: 10.1016/0370-1573(89)90064-1.
[5] S. Freer et al., ‘A single-atom quantum memory in silicon’. arXiv, Sep. 05, 2016. doi: 10.48550/arXiv.1608.07109.
[6] K. J. Morse et al., ‘A photonic platform for donor spin qubits in silicon’, ArXiv160603488 Cond-Mat Physicsquant-Ph, Jun. 2016, Accessed: Jul. 20, 2020. [Online]. Available: http://arxiv.org/abs/1606.03488
[7] C. Chartrand et al., ‘Highly enriched $^{28}$Si reveals remarkable optical linewidths and fine structure for well-known damage centers’, Phys. Rev. B, vol. 98, no. 19, p. 195201, Nov. 2018, doi: 10.1103/PhysRevB.98.195201.
[8] L. Bergeron et al., ‘Characterization of the T center in $^{28}$Si’, ArXiv200608794 Cond-Mat, Jun. 2020, Accessed: Dec. 13, 2020. [Online]. Available: http://arxiv.org/abs/2006.08794
[9] L. Bergeron et al., ‘A silicon-integrated telecom photon-spin interface’, PRX Quantum, vol. 1, no. 2, p. 020301, Oct. 2020, doi: 10.1103/PRXQuantum.1.020301.
[10] A. DeAbreu et al., ‘Waveguide-integrated silicon T centres’. arXiv, Sep. 28, 2022. doi: 10.48550/arXiv.2209.14260.
[11] D. B. Higginbottom et al., ‘Memory and transduction prospects for silicon T centre devices’. arXiv, Sep. 23, 2022. doi: 10.48550/arXiv.2209.11731.
[12] D. B. Higginbottom et al., ‘Optical observation of single spins in silicon’, Nature, vol. 607, no. 7918, pp. 266–270, Jul. 2022, doi: 10.1038/s41586-022-04821-y.
[13] S. Simmons, ‘Scalable Fault-Tolerant Quantum Technologies with Silicon Colour Centres’. arXiv, Nov. 08, 2023. doi: 10.48550/arXiv.2311.04858.
[14] S. D. Barrett and P. Kok, ‘Efficient high-fidelity quantum computation using matter qubits and linear optics’, Phys. Rev. A, vol. 71, no. 6, p. 060310, Jun. 2005, doi: 10.1103/PhysRevA.71.060310.
[15] B. Lackey, ‘Quantum networking: A roadmap to a quantum internet’, Microsoft Azure Quantum Blog. Accessed: Nov. 06, 2023. [Online]. Available: https://cloudblogs.microsoft.com/quantum/2023/11/01/quantum-networking-a-roadmap-to-a-quantum-internet/
[16] GQI, ‘A Deeper Dive Into Microsoft’s Topological Quantum Computer Roadmap’, Quantum Computing Report. Accessed: Nov. 08, 2023. [Online]. Available: https://quantumcomputingreport.com/a-deeper-dive-into-microsofts-topological-quantum-computer-roadmap/
Abbreviations
28Si Silicon 28 isotope
FTQC Fault Tolerant Quantum Computation
GQI Global Quantum Intelligence LLC
MDI-QKD Measurement Device Independent QKD
NISQ Noisy Intermediate Scale Quantum
NIST National Institute of Standards and Technology (US)
PNT Position, Navigation & Timing
PoC Proof of Concept
PQC Post-Quantum Cryptography
QKD Quantum Key Delivery
Q LDPC Quantum Low Density Parity Check
SOI Silicon on Insulator
November 10, 2023