OIDA QUANTUM

PHOTONICS
ROADMAP
Every Photon Counts

March 2020

Created in collaboration with

This OSA Industry Development Associates (OIDA) report is copyrighted
© 2020 OSA Industry Development Associates (OIDA)

---

Please note that this material is copyrighted.

If you reference the material, please provide the complete citation.
If an outside source is cited, please use it.

2
TABLE OF CONTENTS

Executive Summary 4

Acknowledgements 8

The Quantum 2.0 Opportunity: How is quantum technology relevant to the optics and 10
photonics community?
• Context and Objectives for OIDA Roadmap
• Quantum Technology Overview
• Applications and Use Cases
• Global Context: Investments and Activity
• Implications for the Optics and Photonics Community

The Quantum Landscape: Where are the applications leading commercial system 17
development?
• Quantum Sensing & Metrology
• Quantum Communications
• Quantum Computing

Optics and Photonics Component Needs: What are the opportunities for component 37
suppliers?
• Photon Sources
• Photon Detectors
• Optical Fiber / Waveguides
• Modulators & Switches
• Interconnects: Transducers & Converters
• Quantum Memories & Repeaters
• Nonlinear Optics
• Integrated Photonic Platforms

Tracking Progress: What “beacon” milestones or metrics will indicate progress? 55

• Quantum Sensing
• Quantum Communications
• Quantum Computing

Conclusion 59

Bibliography 63

3
EXECUTIVE SUMMARY

Quantum technology is a growing field in physics can be exploited to measure physical properties
and engineering focused on harnessing principles with more precision.
of quantum mechanics to enable functions and
applications not currently achievable with classical- • Communications: Attempts to observe
physics-based technologies. While quantum a quantum communication channel will
mechanics has been studied in the research irreversibly alter the state of the system in a way
community for a century – and has produced that is detectable by the parties exchanging
technologies such as lasers, MRI imagers, transistors, information. A quantum network can distribute
etc. – recent progress in controlling individual entanglement between distant users.
photons, atoms, electrons, etc. has enabled new
advancements. • Computing: Using the principles of superposition
and entanglement, significant speedup over
OIDA developed this roadmap to clarify the market classical computers is theoretically possible for
applications and timing for quantum technologies some problem types.
and to specify improvements in optics and photonics
components needed to enable advancement. Early products are being commercialized today,
The ability to create, manipulate, and read out but quantum technology is still a very new frontier.
states of individual quantum units is expected to Atomic clocks, quantum key distribution systems,
have utility across multiple applications in three and noisy intermediate-scale quantum computers
primary categories: are available in the near-term, but many other
proposed use cases require more advanced
• Sensing and Timing: The extreme sensitivity of hardware and software, and, in some cases, further
quantum systems to environmental influences physics advances.

QUANTUM TECHNOLOGIES TIMELINE -- Illustrative, Not Exhaustive --

2020 2025 2030 2035+

Atomic Clocks GPS-Free Nav.

SENSING SENSING

Gravimeters Medical Imaging

SENSING SENSING

Quantum Random
Number Generators

Quantum Key Distribution Inter- and Intra-City Comms Quantum Internet

COMMUNICATIONS COMMUNICATIONS COMMUNICATIONS

Annealers
COMPUTING

Fault Tolerant Gate-

Noisy Intermediate Scale Gate-Based Quantum Based Quantum
Computer Computer
COMPUTING COMPUTING

*Chevron placement represents anticipated start date of commercialization

Source: Expert interviews, Newry analysis

4
 EXAMPLE APPLICATIONS FOR QUANTUM TECHNOLOGY -- Not Exhaustive --

END MARKETS SENSING & TIMING COMMUNICATIONS COMPUTING

Telecom Clocks for synchronization Cryptography Network optimization

Protecting patient
Medicine Improved brain imaging Drug discovery
data long-term

Protecting critical Drilling location analysis;

Oil & Gas Through-ground imaging
infrastructure oil distribution logistics

Clocks for trade

Finance Secure transactions Portfolio management
timestamping

Battery material
GPS-aided navigation; Cryptography for
Transportation simulation; traffic
quantum LiDAR connected vehicles
optimization

Quantum computing in particular is attracting material science, financial portfolios, distribution

significant attention and funding because of its and logistics, etc. Quantum communications
possible security implications. A universal, fault infrastructure could network individual quantum
tolerant quantum computer would theoretically be computers or sensors together to further enhance
able to factor very large numbers much faster than performance.
a classical computer, which could undermine public
key cryptography methods such as RSA encryption. In light of the opportunities and threats created by
While the industry expects it will take more than a quantum technologies, significant investments are
decade to develop a quantum computer capable being made across the public and private sectors.
of this kind of computation, there is widespread Many governments have large multi-year programs
concern about the potential future vulnerability of with funding levels that exceed US$1 billion. Thirty-
highly sensitive information. two venture capital deals were executed in 2018,
amounting to US$173 million in investment in 2018
This concern is motivating investment in another alone (Gibney, 2019). In addition to venture-backed
realm of quantum technology: quantum key startups, many large, established, multi-national
distribution (QKD). QKD hardware can create a corporations that already supply legacy sensing,
more secure network in which eavesdroppers trying telecommunications, and computing markets
to steal encryption keys would be detected and recognize the potential of and are investing in the
could be circumvented. development and commercialization of quantum
technologies.
Beyond the cybersecurity implications of quantum
technologies, there are many other opportunities In addition to proponents, quantum technologies
for value creation across industries and applications. also have their share of skeptics whose opinions are
Quantum sensors are expected to enable higher- valid. While some applications such as atomic clocks
performance sensors for GPS-free navigation, and point-to-point QKD links are commercially
through-ground imaging, biomedical imaging, available today, current use is modest relative to
etc. Quantum computers are expected to improve their addressable market size due to limitations and
simulation and optimization for drug discovery, tradeoffs between technical performance and cost.

5
Quantum computing has tremendous theoretical will be application-specific, as performance and cost
potential to disrupt and transform a range of advantages need to be demonstrated relative to
industries; however, we are still many years away incumbent sensing approaches. Further integration
from a fault-tolerant quantum computer capable of these systems (e.g., on-chip) would be a notable
of delivering the performance necessary to enable advancement to facilitate lower size, weight,
many applications such as decryption and drug power, and cost (SWAP-C) devices. In quantum
discovery. communications, the development of a quantum
repeater and consistent improvements in distance
For the skeptics who are looking for more evidence and key rates will indicate advancement. Lastly,
of progress, as well as the proponents who wish to implementation of error correction and subsequent
make measured investments that balance risk and scaling in the number of logical qubits will be
uncertainty, we recommend monitoring a few metrics important improvements on the path to developing
and “beacon” milestones that will be indicative of a fault-tolerant quantum computer.
substantial progress. In quantum sensing, progress

/////////////// MILESTONES AND METRICS

Current State Future Vision

Integration.
Gravimeters & Large- Miniaturized
SENSING Format Atomic Clocks Stability, Sensors
SWAP-C

Quantum Repeater,
Point-to-Point Quantum
COMMS Distance,
QKD Links Network
Key Rate

Annealers & NISQ Era Gate-Based Error Correction & Fault-Tolerant

COMPUTING
Computers Logical Qubits Quantum Computer

Advancing Progress: What’s Required

FUNDING HARDWARE
Sustained public and private Performance improvements, cost reduction, supply
investment to support R&D / scaling chain development for components

WORKFORCE SOFTWARE
Education to fill the talent gap and Error correction, algorithm development,
build the future quantum workforce software stack, etc.

6
Optics and photonics are a core enabling element In addition to developing the optics and photonics
of quantum technologies, as many of the systems components for individual subsystems (e.g., sensors,
require very precise control of light. To enable many computing nodes), interconnects to link subsystems
of the applications highlighted above, additional to create a large-scale quantum network or
innovation and supply chain development is computer are also needed. Quantum interconnects
required. Subsequent chapters further specify must transfer the quantum states between various
necessary components to illustrate the many physical media (e.g., atoms, photons, microwave
opportunities available to optics and photonics fields, semiconductor electronics) with high fidelity,
suppliers, including photon sources, photon fast rates, and low loss. The community is beginning
detectors, optical fiber, integrated photonics, to recognize the importance of systems thinking
couplers, modulators, frequency converters, and and engineering for enabling the full ecosystem of
optical-to-microwave transducers. quantum technologies.

SUMMARY: OPTICS AND PHOTONICS COMPONENT REQUIREMENTS Required May Use

Single or Heterodyne and Fiber or
Single Transducers Memories
Entangled Homodyne Integrated
Category Technology Lasers Photon Modulators and or
Photon Photon Photonic
Detectors Converters Repeaters
Sources Detectors Waveguides
Atomic Clocks If networked

Atom
If networked
Sensing Interferometers
NV Center Sensors If networked

Quantum LiDAR
Continuous
Variable QKD
Discrete Variable
Comms.
QKD
Entanglement-
Based QKD
Superconducting If networked

Ion Trap If networked

Neutral Atom If networked

Photonic - Memory-
Discrete Variable dependent
Computing
Photonic - Cont. Memory-
Variable dependent

NV Center If networked

Silicon Spin If networked

Topological TBD TBD

Source: Expert interviews, Newry analysis

7
ACKNOWLEDGEMENTS

Many individuals contributed their perspective and expertise to the development of this report. Although
the following contributors provided many helpful and constructive comments and suggestions, many were
not asked to endorse the conclusions or recommendations of this report and most did not review drafts
before its release.

We extend special acknowledgement to our advisory panel, which was instrumental in scoping our effort
and reviewing and refining the final document:

• Andrew Lord, British Telecom • John Spencer, Photodigm

• Brandon Rodenburg, MITRE • Jun Yang, Corning Incorporated
• Dan Nolan, Corning Incorporated • Michael G. Raymer, University of Oregon
• Gerd Leuchs, Max Planck Institute • Richard Murray, ORCA Computing
• Hugo Zbinden, Université de Genève

Dozens of other individuals also graciously contributed hours of their time to share their perspective and
expertise. We thank and acknowledge all for their contributions:

• Andrew White, University of Queensland • Maithreyi Gopalakrishnan, Quantum Thought

• Anthony Laing, University of Bristol • Mark Day, Alter Technology TUV Nord UK Ltd
• Araceli Venegas-Gomez, QUERCA • Mark Saffman, University of Wisconsin Madison
• Ben Bloom, Atom Computing / ColdQuanta
• Brian Smith, University of Oregon • Martin Schell, Fraunhofer HHI
• Carlos Abellan, Quside Technologies • Michael Fanto, Air Force Research Lab
• Chad Hoyt, Honeywell • Michael Liehr, AIM Photonics
• Dusan Sarenac, University of Waterloo • Mikhail Shalaginov, MIT
• Eleni Diamanti, CNRS • Murray Reed, QLM Technology
• Eric Fossum, Dartmouth College / Gigagjot • Nan Yu, NASA Jet Propulsion Laboratory
Technology • Nils Trautmann, Zeiss
• Fedor Jelezko, Ulm University • Ole Kock, Teledyne
• Franck Pereira dos Santos, LNE-SYTRE • Oliver Slattery, NIST
• James Gates, University of Southampton • Philippe Bouyer, CNRS / Muquans
• Jeremy Levy, Pittsburgh Quantum Institute • Prem Kumar, Northwestern University
• John Berg, PsiQuantum • Ravi Kumar, Atomionics
• John Martinis, Google / UCSB • Sergey Polyakov, NIST
• John Rarity, University of Bristol • Thomas Monz, University of Innsbruck
• Jonathan Olson, Zapata Computing • Tim Spiller, UK Quantum Communications Hub
• Josh Nunn, ORCA Computing / University of • William Munro, NTT
Bath • Yaseera Ismail, University of KwaZulu-Natal
• Karl Berggren, MIT • Zachary Vernon, Xanadu
• Lijun Ma, NIST

8
This work was conducted on an unclassified basis. All assessments of technology-related progress and
feasibility are based on one-on-one conversations with experts, a survey, and other public data. Journal
articles and websites are cited where relevant, but much of the insight was gathered through expert
interviews. The performance targets described in the report are representative of the consensus of a range
of expert interviews.

The OIDA Quantum Photonics Roadmap was developed in collaboration with Corning Incorporated.

The report was researched and written in collaboration with Newry Corp. Newry is a strategy consulting
firm dedicated to fueling growth and enhancing competitive edge for B2B companies in emerging and
technology-intensive markets such as optics and photonics and telecommunications. We help our clients
accelerate decision-making on critical innovation programs, systematically identify growth areas for new
and established products, and anticipate and prepare for future threats and opportunities.

9
THE QUANTUM 2.0 OPPORTUNITY
How is quantum technology relevant to the optics and photonics
community?

Context and Objectives for OIDA Roadmap

Proponents of quantum technologies argue that

quantum systems will transform the way we measure,
process, and transmit information, potentially
disrupting a wide range of end markets from medical
research, to cybersecurity, to telecommunications.

Optics and photonics are key enabling elements

of these transformations; however, there are
uncertainties about what components and
performance levels are specifically required, and
the timing and likelihood of adoption by different
end markets.
Source: Wikimedia Commons

OIDA developed this roadmap to:

Quantum Technology Overview
• Clarify the market potential for quantum
technologies in three technology categories: Quantum technology is a growing field in physics and
computing, communications, and sensing engineering. It is focused on harnessing properties
of quantum mechanics to enable functions and
• Specify what improvements are required in applications not currently achievable with classical
optics and photonics components to enable technology.
advancement

20th CENTURY QUANTUM DISCOVERIES

1905 1985
1935 1964
Einstein’s Laser Cooling
EPR Paradox Bell’s
Photoelectric Experiments Started
Postulated Theorem
Effect

1900 1950 2000

1923
De Broglie
Extended Wave 1935 1957
BCS Theory Formulated
1997
Particle Duality Schrodinger’s
Demonstration
Cat Thought
of C-Not Gate
Experiment
1900
Planck’s Black Body
Radiation Law

10
Quantum mechanics has been studied in the Applications and Use Cases
research community for nearly a century, resulting in
the invention and commercialization of technologies The ability to create, manipulate, and read out
such as laser systems, MRI imagers, transistors, states of individual quantum units is expected to
nuclear power generation, and more. These devices have utility across multiple applications in three
and technologies are viewed as the products of the primary categories:
first quantum revolution, or Quantum 1.0.
• Sensing and Timing: The extreme sensitivity of
As our ability to manipulate and control individual quantum systems to environmental disturbances
quantum objects (e.g., photons, atoms, electrons) can be exploited to measure physical properties
increases, we are approaching a second quantum (e.g., gravity, magnetic fields, time, etc.) with
revolution, or Quantum 2.0, which could further more precision.
enhance performance features such as capacity,
sensitivity, speed, and security for numerous end • Communications: Attempts to eavesdrop
uses. on a quantum communication channel will
irreversibly alter the state of the system, which
Performance breakthroughs are possible due to can be detected by the parties exchanging
unique quantum properties: information and therefore enables more secure
communication.
• Superposition: A quantum system can exist in
two or more basis states at once; its wave function • Computing: Using the principles of superposition
is a linear combination of the contributing basis and entanglement, significant speedup over
states with complex coefficients that reflect the classical computers is theoretically possible
relative weight of each basis state. for some problem types (e.g., factoring large
numbers and certain optimization problems).
• Entanglement: Two or more quantum objects
can be intrinsically linked such that the two Despite high levels of interest from researchers,
states become coupled into an ‘entangled’ the investment community, and the public across
composite state, regardless of how far apart the all three technology categories, Quantum 2.0
objects are from one another. is still a very new frontier. Early products such
as atomic clocks, quantum key distribution
• Measurement: Measurement collapses and systems for quantum cryptography, and noisy
disrupts a quantum state – i.e., the quantum intermediate-scale quantum computers are being
system’s coherent probabilistic superposition commercialized today, but many of the proposed
state collapses into a discrete state, leaving it use cases described on the next page require more
fundamentally changed. advanced hardware and software.

SUPERPOSITION ENTANGLEMENT MEASUREMENT

A quantum object’s Measurement collapses

Two or more
wave function is a the probabilistic
quantum objects can
linear combination of information into a
be intrinsically linked
multiple states discrete classical state

11
 QUANTUM TECHNOLOGIES TIMELINE -- Illustrative, Not Exhaustive --

2020 2025 2030 2035+

Atomic Clocks GPS-Free Nav.

SENSING SENSING

Gravimeters Medical Imaging

SENSING SENSING

Quantum Random
Number Generators

Quantum Key Distribution Inter- and Intra-City Comms Quantum Internet

COMMUNICATIONS COMMUNICATIONS COMMUNICATIONS

Annealers
COMPUTING

Fault Tolerant Gate-

Noisy Intermediate Scale Gate-Based Quantum Based Quantum
Computer Computer
COMPUTING COMPUTING

*Chevron placement represents anticipated start date of commercialization

Source: Expert interviews, Newry analysis

Significant work is still needed to develop these is offered as a major application that is attracting
enabling components and technologies to their significant attention from governments and the
full potential; nevertheless, it is clear that quantum security community. Many public key cryptography
systems could potentially have a major impact methods, such as RSA encryption, rely on the fact
in many end markets, ranging from telecom, to that it is exponentially more difficult for a classical
medicine, to finance, to transportation, and more. computer to factor very large numbers. Quantum
computers, on the other hand, can theoretically
Because we are still in the early stages of quantum factor large numbers very quickly using a quantum
technology development, there most likely will computer algorithm for integer factorization called
be other use cases beyond those mentioned that Shor’s algorithm.
we have not yet imagined. Many observers draw
parallels to the progress that came from the space Many banking transactions, online purchases,
programs of the 20th century, suggesting that critical infrastructure, medical records, and other
investment in quantum technologies will produce applications use RSA or similar encryption methods,
many other enabling component technologies that making them susceptible to a quantum computer
could have utility in other markets, even applications attack. While we are likely 10+ years away from
that we may not be aware of today. having a quantum computer capable of running
Shor’s algorithm, this threat underpins many of the
investments being made globally by governments
Quantum and Data Security and other actors.

A quantum computer’s theoretical ability to factor Threats could still occur in advance of a fault-tolerant
very large numbers faster than a classical computer quantum computer being developed in the form

12
of a “harvest and decrypt” attack. Eavesdroppers In addition to applications where data needs to
may already be tapping fiber networks and storing remain secure for a long period of time, devices that
encrypted data, knowing that they will be able to have long lifecycles and use authenticated over-the-
rapidly decrypt it in the future once a quantum air software updates (e.g., autonomous cars) are
computer is developed. As organizations evaluate also at risk. If a quantum computer is developed, a
the urgency of their data security concerns, they bad actor could potentially get the manufacturer’s
must consider both the “shelf life” of their data (i.e., key and send a malicious software update to hijack
how long it must remain secure) and the time that it the system.
will take to develop, standardize, and migrate to a
new security protocol.

/////////////// EXAMPLE MARKET APPLICATIONS

Clocks for Network

Cryptography
Synchronization Optimization
TELECOM

Improved Brain Protecting Patient

Drug Discovery
Imaging Data Long-Term
MEDICINE

Drilling Location
Through-ground Protecting Critical Analysis; Oil
Imaging Infrastructure Distribution
OIL & GAS Logistics

Clocks for Trade Secure Portfolio

Timestamping Transactions Management
FINANCE

GPS-Aided or GPS- Battery Material

Cryptography for
Free Navigation; Simulation; Traffic
Connected Vehicles
TRANSPORT Quantum LiDAR Optimization

13
 LEVEL OF RISK DETERMINATION
If x + y > z, worry:

Time Until Universal Quantum Computer

z years

Migration Time Security Shelf Life

x years y years

Risk of Secret Keys Revealed

TIME

Source: Mosca, M. (2018, September/October). Cybersecurity in an Era with Quantum Computers: Will We Be Ready? IEEE Security & Privacy, 16(5), 38-41.

Many groups are already working on developing Global Context: Investments and Activity
and evaluating new non-quantum cryptography
methods that may not be susceptible to attack by In light of the opportunities and threats created by
a quantum computer, a field commonly referred to quantum technologies, significant investments are
as post-quantum cryptography (PQC). Additionally, being made worldwide across the public and private
the laws of quantum mechanics can be harnessed sectors. Many governments have large multi-year
to provide an additional layer of security using programs with funding levels that exceed US$1
hardware. Quantum key distribution (QKD) is a billion. Examples include:
quantum communications approach that uses
principles of quantum measurement to create a • USA: The National Quantum Initiative Act was
more secure network. Additional detail is provided signed into public law in late December 2018.
on QKD and PQC in the chapter on Quantum The multi-agency plan involving the DOE, NSF,
Communications. and NIST proposed US$1.2 billion be committed
to funding quantum information science over
While there is a strong sense of urgency around these five years (Giles, 2018). In December 2019 the
data security issues, the timeline for developing a approved federal budget for 2020 included large
fault-tolerant quantum computer capable of running increases (amounting to approximately US$300
Shor’s algorithm is highly uncertain, and leading million) in funding for quantum information
experts admit that they don’t know when – if ever – research support (appropriations specified in
it will be realized. Many compare the development Energy and Water Bill and Commerce, Justice,
required to get from today’s noisy intermediate- Science Bill).
scale quantum computers to a fault-tolerant
quantum computer with the development needed • Canada: Canada has invested over CAN$1
to advance from the transistor to the modern-day billion in quantum R&D over the last decade
integrated circuit. Numerous approaches, each with according to the National Research Council
their respective advantages, are being developed, (Sussman, Corkum, Blais, Cory, & Damascelli,
but it is too early to determine which one(s) will win. 2019).
As one expert put it, “We haven’t had our CMOS
moment yet.” • EU: The European Union’s Quantum Flagship
program is budgeted at €1 billion over a ten-
year timescale (Cartlidge, 2016).

14
• Germany: The German government will invest EXAMPLE GOVERNMENT PROGRAMS
€650 million over two years to support the
transition of quantum technologies from basic US National Quantum Initiative Act
research to market ready applications (IBM, proposes US$1.2 billion over 5 years
2019).

• UK: The UK’s National Quantum Technologies Canada invested over CAN$1 billion over
Programme has exceeded £1 billion in the last decade
cumulative investments since its announcement
in late 2013 (UK Research and Innovation, 2019).
EU Quantum Flagship program budgeted
• China: China is investing heavily in quantum at €1.0B over 10 years
information science. The National Laboratory
for Quantum Information Science received 7B
RMB in funding to start and additional funding China’s National Lab for Quantum Info
is available through national and provincial Science received 7B RMB in funding
initiatives (Kania & Costello, 2018).
UK program has exceeded £1B in
• Japan: Japan set aside more than 30 billion yen cumulative investments since 2013
for quantum applications in a 10-year research
plan (Quantum Business, 2018).
Japan set aside more than 30B yen for
• Australia: Australia’s investment is on the order quantum in a 10-year research plan
of AU$130M through federal funding alone
(Roberson & White, 2019).

Private sector investment is also increasing worldwide. Many startups are spinning out of academic research
across sensing, communications, and computing applications. Additionally, large, established multi-
national corporations that already supply legacy markets are recognizing the potential impact of quantum
technology and are investing in its development and commercialization. The number and value of venture
capital deals, particularly in quantum software and computing startups, is increasing, reaching a total of 32
deals in 2018 at a total value of US$173 million in 2018 (Gibney, 2019).

Given that we are likely many years away from a fault-tolerant quantum computer or a quantum internet, it
is possible that the investment community will lose interest if the technology does not advance fast enough
to deliver on expectations. Some express concerns that a “Quantum Winter” – i.e., a freeze in venture
capital and/or government funding – might be coming.

EXAMPLE COMPANIES INVOLVED IN QUANTUM TECHNOLOGIES

SENSING & TIMING COMMUNICATIONS COMPUTING

AO Sense Microsemi ArQit Electric Atom IonQ
Atomionics Muquans Fujitsu NEC Computing Microsoft
Bosch Rydberg Huawei NTT ColdQuanta ORCA
Honeywell Technologies ID Quantique QuantumCTek Google Computing
ID Quantique Teledyne e2v Infiniquant Qubitekk Honeywell Psi Quantum
Lockheed Thales KETS Quintessence IBM Rigetti
Martin Vector Atomic Mitsubishi Labs Intel Xanadu

15
 NUMBER OF VC DEALS

35 $300M
Instrumentation, tools, and services 32
30 Communication 3
$250M
Computing
25 7
Software $200M
20 Sensors and materials
17 6
Total value of deals (Millions of USD) 1 $150M
15 1
12
1 $100M
10 9 9 10
2 14
4 7
5 4 $50M
3 2
1 2 5
3 3 2 2
0 $0M
2012 2013 2014 2015 2016 2017 2018
Source: Adapted with permission from Springer Nature: Gibney, E. (2019). The quantum gold rush: the private funding pouring into quantum startups.
Nature, 574, 22-24.

Some investors are aware of this potential but asserted that “a quantum arms race is more likely
are hopeful that incremental value created by than a quantum winter.”
near-term technologies (e.g., quantum sensors,
QKD networks, noisy intermediate-scale quantum
computers) will be sufficient to justify the cost of Implications for the Optics and Photonics
ongoing development efforts. Other investors are Community
hoping that enough progress will be made that a
different firm in a later-stage deal will buy them out. Optics and photonics are a core enabling element
A minority are betting that a breakthrough might of quantum technologies, as many of the systems
occur within the next five years to dramatically require very precise control of light (Awschalom, D.,
accelerate the timeline. Berggren, K. K., Bernien, et. al., 2019). Quantum
technology presents various innovation opportunities
Irrespective of the hardware advancements required, for the optics and photonics community, and there
significant parallel software and application is a large ecosystem of component, system, and
development is ongoing. Many potential end users service players with various requirements for their
such as chemical and energy companies are actively respective architectures.
engaging with quantum software development
companies to evaluate what problems could be Furthermore, system developers have consistently
solved by a quantum computer once the hardware expressed a need for a more robust and developed
is more advanced. supply chain. Subsequent chapters further specify
necessary components to illustrate the many
While private funding trends are likely to depend opportunities available to optics and photonics
on progress demonstrated in the near term, suppliers to participate in this nascent but growing
the large, multi-year government programs are market.
expected to subsidize progress even if private
funding levels decline. Given the potential national In addition to opportunities, all entities (not just
security implications for quantum technology (e.g., optics suppliers) should be aware of the potential
decoding encrypted messages, detection of stealth threats that quantum computers pose to their
aircraft with quantum radar, etc.), one expert even information security.

16
THE QUANTUM LANDSCAPE
What are the applications leading commercial system development?

INTRODUCTION TO QUANTUM SENSING AND TIMING

While quantum computing receives most of the attention by the mainstream media, most experts
believe that quantum sensors and clocks will bring earlier commercial success if they can outperform
conventional techniques cost effectively and in a smaller and often portable form factor. In fact,
some types of quantum sensors – such as atomic clocks and cold atom interferometers – are already
commercially available. Work is ongoing to bring new products to market, while making existing
systems smaller and more robust.

Quantum sensors take advantage of the extraordinary sensitivity of quantum states to their environment
to create more sensitive and precise measurement devices to improve navigation (with or without
GPS), network synchronization, geological surveying, medical imaging, LiDAR, etc. There are multiple
technological approaches to building high-performance quantum sensors, which are outlined below.

Atomic Clocks

Atomic clocks use the hyperfine transitions in atoms should have nearly constant and identical tick rates
as a frequency reference from which they derive and they should experience minimal drift rate over
a time standard. The high frequency of atomic time (uncertainty levels as low as 10-18).
transitions enables extremely high-resolution
timing. Furthermore, while man-made quartz-crystal Lasers, heterodyne and homodyne photon
oscillators used in traditional clocks vary slightly detectors, and modulators are common optics and
from one to the next, all atoms of a given element photonics components required for atomic clocks.
and isotope are identical. Thus, as long as the Optical fiber, integrated photonic waveguides, and
atoms are properly isolated from outside influences, transducers and converters may also be used.
separate atomic clocks based on the same isotope

QUANTUM SENSOR COMMERCIALIZATION TIMELINE -- Illustrative, Not Exhaustive --

2020 2025 2030 2035+

Atomic Clocks
SENSING

Gravimeters & Gradiometers GPS-Free Nav.

SENSING SENSING

Magnetometers for Brain Imaging

SENSING

Quantum LiDAR
SENSING

*Chevron placement represents anticipated start date of commercialization

Source: Expert interviews, Newry analysis

17
There are many applications that could benefit from For example, local gravity will impart a phase shift
the performance improvements provided by atomic on an atomic matter wave because it has mass,
clocks. Examples include: which modifies the interference pattern in a way that
can be detected to create a gravity sensor. Cold-
• GPS: Precise time measurement is critical for atom systems isolate a cloud of atoms (e.g., Rb) in
global positioning systems to ensure accurate a vacuum and manipulate the states of these atoms
determination of the distance between satellites using lasers. A laser can put each atom into a hybrid
and a ground receiver. (superposition) state so that it simultaneously has
slightly lower and higher energies. Then, a laser will
• Network Synchronization / Time Stamping: give the atom cloud a kick so that the higher-energy
Accurate time coordination and time stamping half of the atom’s wavefunction moves up by a few
is essential for synchronizing and sequencing centimeters from the low-energy half.
operations or transactions in a network, be it
a power grid, telecommunications network, air Next, one more kick moves the two wavefunction
traffic control, financial systems, etc. halves back together, creating an interference
pattern in the presence of a third kick. The
• Research: Dark matter research and other gravitational potential energy of the cloud affects
fundamental science experiments can benefit the wavefunction, so the interference pattern
from the precision made possible by atomic can be analyzed to determine the strength of the
clocks. gravitational field. Gradiometers may use the same
principle as gravimeters, but with two clouds at
Atomic clocks were commercialized decades different heights to determine the gravitational
ago and multiple companies such as AOSense, field gradient.
Muquans, Vector Atomic, Frequency Electronics,
ADVA, Microchip Technology, and Teledyne are -- Illustrative --
developing and commercializing more advanced Kicked atoms
systems. Current research and engineering efforts
are focused on increasing environmental robustness
and stability and miniaturizing the systems to be
Gravity

lower size, weight, power, and cost (SWAP-C) to

allow for deployment in higher volume, but more
cost-sensitive and/or mobile applications.
Unkicked atoms
Atom Interferometers for Gravity, Inertia, and Detector
Rotation Sensors
Kicked and unkicked atoms will follow
Cold atom interferometers can be applied to create slightly different paths and create an
faster and more sensitive devices for measuring interference pattern on the detector
gravity, inertia, and rotation. They measure the Source: Newry
interference pattern between atomic matter waves
traveling along different paths, which allows them As described, lasers, coherent photon detectors,
to measure a variety of physical signals (e.g., speed, and modulators are required. Waveguides,
acceleration, etc.). Like atomic clocks, they are also transducers, and converters may also be needed
less susceptible to drift than conventional methods. depending on the application.

18
High-precision, drift-free atom interferometers can 1. Cold atoms
be used in several applications, such as: 2. Nitrogen-vacancy (NV) centers in synthetic
diamond crystals
• Climate Research: Satellites equipped with a
gravimeter may be used to measure gravity Cold atom-based magnetometers isolate a vapor of
across different regions of the earth to inform atoms (e.g., Cs, Rb) in a vacuum chamber. A laser
climate models of polar ice mass, ocean currents, beam is applied to orient the valence electron spins
and sea level, which in turn can aid researchers’ of all atoms in the same direction. Any external
predictions of natural disasters (e.g., floods, magnetic field will cause the spins to wobble in
earthquakes, volcano eruptions). a motion called precession. Precession causes a
change in the laser light, which can be measured
• Civil Engineering: More precise quantum to determine the strength of the external magnetic
gradiometers could better identify pipes, field.
mineshafts, tunnels, and sinkholes in or near
work sites, resulting in improved safety for Nitrogen-vacancy-based magnetometers rely on
workers and significant cost savings via more the behavior of point defects in synthetic diamond
accurate digging. crystals to detect the strength of magnetic fields.
NV centers – nitrogen atoms located in vacancies in
• Hydrocarbon and Mineral Exploration: Any the diamond lattice – emit red light (637 nm) when
object or cavity with a lower density than its excited by green light (532 nm).
surroundings will have a lower local force of
gravity. Quantum gravimeters can measure and Due to spin-dependent decay pathways, the
map these dynamics to guide natural resource fluorescence intensity of these emissions varies
investigation. based on the spin state of the NV center, which is
affected by the local magnetic field. The strength of
• GPS-Free Navigation: Extremely accurate, that magnetic field can be determined by measuring
precise, and stable quantum accelerometers the variance in fluorescence intensity with microwave
and gyroscopes based on atom interferometry and laser pulses. NV centers are also sensitive to
would not need to rely on GPS satellites, offering other external signals and may be used to measure
a complementary alternative for navigation electric field, strain, and temperature.
in environments that do not have access to
satellites or in applications that are susceptible Quantum magnetometers are being developed for
to a GPS jamming or spoofing attack. Map various end applications:
matching with gravity or magnetic sensors may
also be used for GPS-free navigation. • Brain Imaging: Highly sensitive quantum
magnetometers can measure the magnetic fields
of electric currents in the brain (often referred to
Magnetometers as magnetoencephalography) to study epilepsy,
Parkinson’s disease, dementia, etc.
Highly sensitive magnetic-field sensors, called
magnetometers, have the potential to improve • Heart Imaging: Measuring magnetic fields from
navigation and various imaging applications in induced electric currents can be used to identify
arenas such as healthcare and manufacturing. sources of atrial fibrillation (technique called
Magnetometers can be developed using one of magnetic induction tomography); quantum
two approaches: sensors have the potential to offer lower
SWAP-C than incumbent instruments.

19
• Metrology: Magnetometers may be used Quantum LiDAR is still being developed but may
for defect analysis in nanoelectronic circuits, leverage the single-photon detectors developed for
detection of micro-fissures, etc. the quantum communications market. Proposed use
cases include autonomous vehicles and quantitative
• Navigation: Extremely sensitive magnetometers imaging for leak detection for methane gas and
may also be used to measure the direction and carbon dioxide. Companies like QLM Technology
strength of magnetic field anomalies. Location and ID Quantique are commercializing products for
can be determined by overlaying the local the leak detection application.
measurement data on reference maps of the
Earth’s magnetic field. Commercialization of quantum LiDAR for gas and
leak detection is expected to occur within the
Quantum magnetometers, particularly NV center- next 1-3 years, whereas automotive applications
based magnetometers, are at an earlier stage are longer term. In addition to performance
than the atomic clocks and cold-atom gravimeters improvements and miniaturization, significant cost
described previously. Consistent fabrication of reductions will be required to make the technology
NV centers remains a challenge, and further feasible for high-volume, cost-sensitive automotive
improvements in sensitivity are needed for many markets.
applications.
There is limited public information about the
Academic researchers and startups, including commercialization status of quantum radar and
Nvision Imaging Technologies, QuSpin, and there is some debate as to whether it is possible
Quantum Diamond Technologies, continue to at all. There have been announcements that China
invest in advancing the technology. Large multi- has developed a quantum radar system, but no
national corporations (e.g., Bosch, Thales, Zeiss, data supporting this claim has been shared. Most
Lockheed Martin) are working on the technology as experts agree the commercialization timeline for
well, but no product has been commercialized yet. quantum radar is longer term than the other sensors
While promising results are being published today, described previously.
additional engineering and the regulatory approval
process is expected to delay commercialization of
this technology for medical applications. Imaging

Quantum-enhanced imaging is another field that

Quantum LiDAR/Radar harnesses the properties of non-classical states of
light to enable improved imaging performance (e.g.,
Quantum LiDAR may use single-photon detectors resolution) (Zeiss, 2018). Examples include ghost
and in some cases entangled light sources to enable imaging (improves signal-to-noise for applications
more sensitive LiDAR systems capable of operating with low intensity), quantum multiphoton
in poor visibility environments. Quantum radar microscopy, quantum coherence tomography,
has also been theorized and is reportedly being quantum interferometry, quantum lithography,
developed. The scheme involves bouncing one half etc. Quantum enhanced imaging has relatively less
of a series of entangled photon pairs off an object commercial activity compared to other quantum
and then comparing the returning photons with sensing modalities but may likely find use in niche
the idler photons that were held back. This method applications in the next several years.
could distinguish the radiation of the entangled
photons from other sources of noise to spot stealth
aircraft, for example. Higher-resolution images are
also expected with this approach.

20
 INTRODUCTION TO QUANTUM COMMUNICATIONS

Quantum communications leverages the properties of quantum state preparation and measurement
as well as intrinsic quantum phenomena such as entanglement and squeezing to create secure
communications networks and to enable a range of longer-term applications including distributed
sensing and remote or distributed quantum computing.

A primary motivation for developing quantum communications networks is to create secure

communication networks to protect against the potential data security threat posed by quantum
computing. A quantum computer capable of implementing Shor’s algorithm could factor large integers
exponentially faster than a classical computer, rendering common asymmetric public key encryption
protocols such as RSA ineffective (National Academies of Sciences, Engineering, and Medicine, 2019).

RSA is currently used to encrypt and distribute encryption keys, which are then used to encrypt the
data that is subsequently transmitted. New key encryption methods that are not susceptible to a
quantum computing attack are needed, even in advance of the development of a fault-tolerant
quantum computer. Encrypted data that is being exchanged today and needs remain secure long-
term is at risk of being stored now and decrypted later once large fault-tolerant quantum computer
is developed.

While we are likely a decade or more away from a quantum computer capable of breaking RSA
2048, the industry is actively working on new solutions to resist quantum computing attacks, and
some applications and end users are already migrating to quantum-safe solutions today. In the
software-based realm of post-quantum cryptography (PQC), quantum-resistant algorithms are being
developed. Quantum key distribution (QKD) is a complementary hardware-based approach that is
already commercially available and further development is ongoing. While solution selection will likely
be application-dependent, most cryptography experts agree that a hybrid approach will be used in
the highest security applications for defense in depth.

QUANTUM COMMUNICATIONS COMMERCIALIZATION TIMELINE -- Illustrative, Not Exhaustive --

2020 2025 2030 2035+

Fiber-Based QKD Quantum Internet

COMMUNICATIONS COMMUNICATIONS

Satellite-Based QKD
COMMUNICATIONS

QRNGs
COMMUNICATIONS

Post-Quantum Cryptography
COMMUNICATIONS

*Chevron placement represents anticipated start date of commercialization

Source: Expert interviews, Newry analysis

21
 POTENTIAL SOLUTIONS
If encrypted data is
stolen, a quantum Post Quantum Cryptography (PQC)
computer capable of
Algorithms that (unlike RSA and ECC) are quantum-
running Shor’s algorithm
resistant – i.e., encryption methods based on math that
could crack some public
a quantum computer is not advantaged in computing
key encryption protocols
such as RSA much faster
than a classical computer, Quantum Key Distribution (QKD)
rendering them insecure. Hardware-based approach that facilitates key exchange
New “quantum secure” by exchanging photons which, by the principles of
key exchange solutions quantum physics, will be perturbed in a detectable way
are required. if an eavesdropper is present

Source: Newry

Post-Quantum Cryptography standards are expected to be published around

2022.
Post-quantum cryptography solutions, also
described as quantum-proof or quantum-resistant,
are cryptography techniques that are expected to Quantum Key Distribution Approaches
be secure against known attacks from a quantum
computer. These algorithmic solutions are purely Quantum key distribution is an alternative but
based on new mathematical models and do not complementary hardware-based approach to PQC.
incorporate any quantum physical systems; rather, It involves exchanging an encryption key between
they would be a replacement for RSA to secure key parties over a quantum network, which then enables
exchange. secure encrypted communication via the classical
internet.
Symmetric keys have been proposed as one of
many possible solutions and protocols are already QKD systems make use of quantum random
available. While a quantum computer running number generators (QRNGs) to create truly random
Grover’s algorithm could potentially diminish the keys by leveraging the inherent randomness in
effective security of symmetric keys, increasing the quantum systems. In addition to their applicability
key length is expected to render them more secure. in cryptography, QRNGs also have utility in
Many new protocols are being proposed and applications such as electronic gaming, numerical
evaluated, leveraging a multitude of mathematical simulations, and the Internet of Things (IoT).
approaches such as lattice-based, code-based, and QRNGs are already commercially available through
multivariate. suppliers such as IDQuantique, Quintessence Labs,
Quside, etc.
Cybersecurity firms such as ISARA Corporation
are already providing recommendations to clients There are multiple protocols and hardware
about quantum-safe cryptography options. NIST approaches for transmitting these keys, including
has a large PQC standardization effort underway Prepare and Measure QKD and entanglement-
focused on evaluating alternative approaches for based QKD. Entanglement-based QKD has not
public key encryption and digital signatures. Draft been deployed commercially yet.

22
Prepare and Measure QKD Entanglement-Based QKD

-- Illustrative -- -- Illustrative --
Photons
Alice Bob Alice Bob

Entanglement
Photons
Alice Bob Alice Bob

Entanglement
Eve Eve
Source: Newry Source: Newry

Prepare and Measure QKD protocols take If Alice and Bob establish an entangled photon pair,
advantage of quantum measurement – i.e., the fact it can be used as the qubit to hold the quantum
that measuring a qubit collapses its probabilistic information necessary to generate a secrete key
wave function into a discrete state. later on. Alice and Bob may measure directly
on the photons carrying the entanglement. The
Consider the case of Alice attempting to transmit a information about the measurements can then
secret encryption key to Bob. Alice, the transmitter be shared via the classical internet to negotiate a
prepares photons in certain states in a certain ‘state secret shared key.
basis’ according to randomly generated numbers
and sends them over the quantum internet to As with other QKD approaches, this scheme
Bob. Bob measures the photons to determine the is secure because if Eve tries to eavesdrop, it
prepared state. will cause a detectable change in the system,
enabling Alice and Bob to determine whether the
If an eavesdropper, Eve, attempts to measure the exchanged key is compromised before using it.
photons along the transmission path between Eve
and Bob, she will alter their state in a detectable This scheme is also device independent, meaning
way. Thus, Alice and Bob can determine whether that it is secure even if the devices used to
the key has been compromised before using it distribute the keys between Alice and Bob have
to communicate securely. And, if the disturbance been maliciously prepared by Eve.
introduced by Eve is below a known threshold,
Alice and Bob can use classical error correction and
privacy amplification to still obtain a secure key,
although reduced in length.

There are two types of Prepare and Measure QKD

schemes that leverage different detectors:

• Discrete Variable QKD (DV-QKD): Uses single-

photon detectors.

• Continuous Variable QKD (CV-QKD): Uses

homodyne detection / field quadratures.

23
 Source: Khan, I., Heim, B., Neuzner, A., & Marquardt, C. (2018, February). Satellite-Based QKD. Optics & Photonics News, 26-33.

Quantum Key Distribution Hardware

Each of the described approaches uses different components.

Photon Sources count rate, and cooling requirements. There are also
major tradeoffs between performance and cost for
Modulated weak (low-power) lasers, single-photon single photon detectors. While SNSPDs are highest
sources, or entangled photon sources are needed performance, they require cryogenic cooling, which
for QKD. Polarization-modulated single photons are makes them cost prohibitive for some applications.
an intuitive choice for a qubit but are challenging to
implement in practice due to current performance Fiber-Based vs. Satellite-Based QKD
levels of single-photon sources. As an alternative,
many protocols may use weak laser pulses, which Implementation of these protocols can use either
are already commercially available, that may be fiber-optic or free-space (satellite-based) links.
modulated in phase, amplitude, or polarization. QKD has already been installed in existing fiber
optic communication networks but is limited in
Both single-photon and laser-based schemes its transmission distance as well as its rate. QKD
assume that the receiver trusts the sender; the networks cannot use traditional, classical repeaters
sender would have a memory where the key is because amplification would alter the quantum state
stored. Entangled photon sources, while the of the transmitted photons. Current installations are
least developed, are the only sources that don’t limited to distances of less than a few hundreds of
necessitate that intrinsic trust. kilometers because of this constraint.

Photon Detectors To overcome this distance limitation, many possible

solutions are being worked on:
Single-photon detectors (avalanche photodiodes
[APDs] or superconducting nanowire single-photon • Development of quantum repeaters: Many
detectors [SNSPDs]) or, in the case of continuous research groups are working to develop
variables, homodyne detectors (PIN diode-based) quantum repeaters that would be able to
may be used to receive the transmitted photons. store, manipulate, and re-send qubits without
Different protocols (e.g., discrete vs. continuous disrupting their quantum state; however, it will
variable) and applications will use different detectors likely be several years before a quantum repeater
to optimize for tradeoffs such as efficiency, dark is commercially realized.

24
 • Using trusted nodes: Some installations are using The principle has already been demonstrated: China
classical nodes to refresh the signal, piecing famously launched the first publicly known QKD
together multiple independent, shorter QKD satellite, Micius, in 2016, creating an encrypted
links. Keys are established separately for each intercontinental video link to another research
section. These nodes are typically called trusted group in Vienna (Khan, Heim, Neuzner, & Marquardt,
nodes because they require security at each link 2018). The technology is currently limited to low
to ensure the information is not compromised. transmission rates and night-time operation, but
Some argue this is only a short-term solution development is ongoing. ArQit, a London-based
because there is still potential for compromised startup, is working to build the first commercial
security while others believe it could be a satellite-based QKD system. There is also an EU-
practical approach for some applications if the based QUARTZ Satellite Cybersecurity Consortium.
costs of the QKD transmitter/receiver boxes fall. This approach will present additional requirements
Nevertheless, trusted-node quantum networks for optics and photonics components, as they will
have already been demonstrated. China, for need to be rugged enough for use in space.
example, has a 2000 km network connecting
Beijing and Shanghai with 32 trusted nodes QKD hardware is already commercially available from
(Courtland, 2016). suppliers such as ID Quantique and Quintessence
Labs and has been installed in various networks.
• Using satellites and free-space optical links: Many other large corporations have active research
Propagation losses scale exponentially with groups as well, including Toshiba, NTT, NEC, Fujitsu,
distance in fiber, but only quadratically in free Mitsubishi Electric, etc.
space, making space-based QKD an attractive
alternative to fiber-based QKD (Khan, Heim, Networks and trials are already installed for various
Neuzner, & Marquardt, 2018). Free-space applications. For example:
optical communication can also operate over
a larger range of wavelengths relative to fiber- • Financial Networks: In Fall 2018, several major
based systems, which must operate in the banks and financial institutions began testing a
telecommunications band (Awschalom, D., QKD network between Manhattan and northern
Berggren, K. K., Bernien, et. al., 2019). This New Jersey via the Holland Tunnel (Kahn, 2019).
method would use low-earth orbit satellites to
distribute secure keys to ground stations via • Telecom: SK Telecom announced they plan to
free-space optical links. install QKD technology on the Seoul-Daejeon
section of its LTE and 5G networks (SK Telecom,
2019).

• Government: In 2007, QKD was trialed to protect

the Geneva state elections in Switzerland (ID
Quantique, 2017).

Other applications that are also expected to

Vienna Beijing adopt include corporate networks, healthcare,
defense, etc.; however, additional improvements
Vienna Beijing are needed, including increased range, improved
key rates, and lower costs / miniaturized hardware.
Quantum memories and repeaters will be vital
Source: Newry for increasing the reach of fiber-based QKD
networks. Active stabilization and synchronization
of components at the transmitters and receivers

25
(such as laser frequency, polarization, phase of two quantum sensing networks. These end nodes may
optical interferometers) is also critical for long-term consist of simple photonic devices or small quantum
operation. Standards also need to be developed computers.
and are underway through initiatives at the ITU and
ETSI. In addition to hardware, software stacks, network
stacks, and control planes need to be developed
From a system engineering perspective, further to enable the quantum internet. Key milestones to
integration and compatibility with existing monitor progress toward a functioning quantum
infrastructure, networks, and value chain players is internet are outlined in a subsequent chapter.
another major hurdle. While protocols have been Incremental progress is being made today, but
developed, additional experimentation and test many experts expect it will take at least another
beds are required to prove their effectiveness, decade to develop a long-distance, entanglement-
which will take time. These testbeds are being based network.  
developed; for example, Europe’s OPENQKD
network project is being used to demonstrate
integration of QKD-enabled technologies to help
advanced standardization and certification efforts.

Beyond QKD: Quantum Network/Internet

While QKD is expected to be the first application

for quantum communications, a quantum internet
with more functionality is expected to be developed
longer term, which would enable distributed
networks of quantum sensors and quantum
computers. While QKD can enable more secure
communication, a quantum internet could enable
other applications such as:

• Distributed network of quantum sensors to

increase performance

• Secure quantum computing in the cloud (i.e.,

blind quantum computing)

• Quantum computing clusters – distributed

quantum computing

Additional development is required to realize these

applications. For example, quantum memories are
required to enable cloud quantum computing so
that two users can obtain and store entangled qubits
and use them to teleport quantum information to
one another. End nodes for preparing, measuring,
storing, and manipulating qubits will also be
required for distributed quantum computing and

26
 INTRODUCTION TO QUANTUM COMPUTING

By harnessing superposition and entanglement, quantum computers have the potential to solve
certain types of problems exponentially faster than classical computers. Where classical computers
carry out logical operations on binary bits which can be 0 or 1, quantum computers’ bits, called
qubits, can be in a superposition state of 0 and 1, allowing for a much larger computational space.

Quantum computers will work in a hybrid architecture with classical computing in most cases.
Research is still ongoing to identify the range of problems for which quantum computing can truly
provide an advantage and it is very unlikely that quantum computers will completely replace classical
computers in the foreseeable future. Most believe that quantum computers are more likely to be
specialized machines applied to specific types of problems they are uniquely suited to solve.

COMPARING QUANTUM VS. CLASSICAL

Demonstrating that a quantum computer can solve a problem

Quantum Advantage (irrespective of the usefulness of the problem) faster or lower cost than
a classical computer

Demonstrating that a quantum computer can solve a problem, again

Quantum Supremacy irrespective of the usefulness of the problem, that a classical computer
cannot solve in a practical length of time

Demonstrating quantum advantage or supremacy for a problem that

Quantum Value has real commercial relevance where value (e.g., a better solution,
shorter run time) is created
Source: Newry

Quantum Computing Applications In several applications where quantum computers

are expected to have an advantage, some algorithms
Quantum computing could have broad applicability, have already been developed, and work is ongoing
although experts in the community are wary of over- across many end markets to test the utility of early
hyping the technology. They caution that for each machines on real-world problems. In other cases,
proposed application, it is critical to understand: additional algorithms need to be developed and
simultaneous improvements in the hardware
• What algorithm will be used? (e.g., more qubits, lower error rates, etc.) will be
required to implement them. Some firms are even
• What is the scale of the advantage realized with developing software in advance of the hardware
a quantum computer compared to a classical being available. Quantum computing is expected
computer? to be particularly well-suited for the problem types
outlined in more detail below.

27
Large Number Factoring More precise classification of data according to
different features is expected to improve the
As described in the Quantum Communications performance of artificial intelligence systems.
chapter, quantum computers are theoretically Multiple algorithms may be applied to solve these
capable of leveraging Shor’s algorithm to crack problems more effectively, including quantum
common public-key cryptography methods. First amplitude estimator, quantum support vector
devised in 1994, roughly a decade after quantum machines, and Harrow, Hassidim, and Lloyd’s
computers were initially proposed, Shor’s algorithm algorithms.
revealed a new approach to factoring large numbers
and reignited interest in developing a quantum Optimization
computer.
Quantum computers can be applied to evaluate
Simulation many potential solutions extremely quickly, which
could apply to vehicle routing for distribution,
The automotive and pharmaceutical industries manufacturing process planning and supply chain
are already exploring the advantages of quantum optimization, financial portfolio management,
computers for chemical simulation for molecular network optimization, etc. End users already trialing
design of new materials and drugs. For example, these examples include Airbus (evaluating the most
VW is trialing early quantum computing machines to fuel-efficient ascent paths), VW (traffic flow routing),
simulate various material options for electric vehicle and British Telecom (cell network optimization)
batteries (Volkswagen, 2018). Pharmaceutical (D-Wave Systems Inc., 2019).
companies, such as Biogen, are evaluating the utility
of quantum computers for their drug discovery Database Searching
process (Accenture, 2020).
Grover’s algorithm is another famous quantum
Quantum computers could also provide an algorithm that can be applied to search large,
advantage for scenario simulation in end markets unsorted databases. It is expected to deliver
spanning distribution and logistics (e.g., scheduling), quadratic speedup, compared to the exponential
financial services, and healthcare and life sciences speedup expected from Shor’s algorithm.
by simulating multiple potential scenarios and their
associated outcomes. The increased computing
power could improve assessments of risk, pricing,
and other market dynamics. Quantum amplitude
estimator algorithms may be applied for this
purpose.

Machine Leaning and Artificial Intelligence

Quantum computing has been proposed to

improve artificial intelligence and machine learning
applications by finding relationships and patterns in
data (Carleo, et al., 2019). For example, MIT and IBM
have collaborated to assess the advantage of using
quantum computers for classification, especially
feature mapping (Temme & Gambetta, 2019).

28
IBM ROADMAP: QUANTUM COMPUTING USE CASES
Health Care
Chemicals and Distribution and Financial
and Life Manufacturing
Petroleum Logistics Services
Sciences
Chemical product Materials
Drug discovery
Chemical design discovery
Simulation Protein structure
Surfactants, Quantum
predictions
catalysts chemistry
Derivatives pricing
Scenario Disruption Disease risk
Simulation management Investment risk predictions
analysis
Feedstock to Distribution supply Fabrication
product chain Portfolio optimization
management Medical / drug
Optimization Oil shipping / Network
supply chain
Manufacturing
trucking optimization Transaction supply chain
settlement
Refining processes Vehicle routing Process planning
Finance offer
Consumer offer Accelerate
recommender Quality control
recommender diagnosis
Drilling locations Credit / asset
AI / ML Freight forecasting Genomic analysis Structural design
Seismic imaging scoring and fluid
Irregular behaviors Clinical trial dynamics
Irregular behaviors
(ops) enhancements
(fraud)
Source: IBM, Accessed on https://quantumcomputingreport.com/our-take/a-quantum-computing-application-roadmap-from-ibm/

IBM ROADMAP: QUANTUM COMPUTING ALGORITHMS

Variational Quantum Uses energy states to calculate the function of the variables to optimize; efficiently
Eigensolver calculates complex portion of simulations, but requires many qubits for large problems

Quantum
Optimizes combinatorial style problems to find solutions with complex constraints;
Approximate
enables robust optimization in complex scenarios, but need to expand to more classes
Optimization

Quantum Amplitude Creates simulation scenarios by estimating an unknown property; potential to speed up
Estimator simulations to solve dynamic problems, but high quantum volume needed for efficiency

Quantum Support Supervised machine learning for dimensional problem sets; allows for better separation
Vector Machines of data points to improve accuracy, but runtime can be slowed by data structure

Harrow, Hassidim, Estimates the measurement of large linear systems; could enable exponential speedup
and Lloyd of math calculations, but challenge is it’s hard to satisfy prerequisites

Source: IBM, Accessed on https://quantumcomputingreport.com/our-take/a-quantum-computing-application-roadmap-from-ibm/

29
Types of Quantum Computing Quantum annealers and simulators use a different
approach to quantum research than gate-model
There are two general classes of quantum machines and are well-suited for optimization and
computers: analog and gate-based. Analog systems simulation problems. There are questions about
include simulated annealers, adiabatic computers, whether annealing machines will be able to run
and more general-purpose quantum simulators, all Shor’s algorithm directly. D-Wave has already
of which directly manipulate interactions between developed commercialized annealers that are being
physical qubits rather than breaking actions into used by a range of customers, including Los Alamos
more abstract gate operations. Analog machines National Laboratories, Volkswagen and OTI. These
involve a system with qubits prepared in an initial customers have either installed systems or have
quantum state; then, the Hamiltonian is adiabatically access to D-Wave’s cloud platform, Leap. Many
changed to the final Hamiltonian such that the applications are being tested, and development
problem solution (answer) corresponds to the final efforts are ongoing to further scale these systems.
state of a modified Hamiltonian. Because the value
of the qubits are analog, it is accordingly difficult Gate-based quantum computers are currently at
to implement error correction because the set of an earlier stage. While early systems have been
quantum operators are nondiscrete. developed and cloud access is even possible for
some, they are still relatively small-scale and error
In contrast, gate-based quantum computers, prone. One key challenge for all quantum computers
sometimes referred to as universal quantum is that qubits tend to lose their superposition or
computers, use logical gate operations (AND, OR, coherence when they interact with the environment
etc.) on physical qubits, as their name suggests. – a phenomenon called decoherence. Environmental
Each gate operation is performed by allowing noise (e.g., vibrations, temperature changes) can
qubit evolution for a certain duration driven collapse the qubit into the ground (or a thermally
by a Hamiltonian that can instigate the desired excited state) quantum state before the calculation
transformation of qubits. Quantum error correction is finished, which is why many of these systems
is possible as long as there are a sufficient number require vacuum and cryogenic systems to isolate
of qubits available to provide redundancy. the qubits.

EXAMPLE D-WAVE CUSTOMERS AND APPLICATION CATEGORIES

Los Alamos Optimization, machine learning, OTI Materials science

National Lab simulating quantum systems

Traffic flow optimization, battery

British Cell phone network
Volkswagen simulation, acoustic shape
Telecom optimization
optimization

Air traffic route optimization, Machine learning for election

DLR airport gate scheduling Qx Branch modeling

Traffic flow optimization,

Financial portfolio
manufacturing process
Denso Nomura optimization
optimization

Source: Media courtesy of D-Wave Systems

30
 GATE-BASED QUANTUM COMPUTING APPROACHES

Superconducting Spin Topological Ion Trap Neutral Atom Photonic

Uses the spin Ions trapped in Atoms trapped in Qubits encoded

Superpositions of Maintain quantum
state of electric fields in magnetic or optical in the quantum
currents flowing in states in large clouds
electrons vacuum fields in vacuum states of
opposite directions of electrons rather
confined in chambers – uses chambers – uses photons
around a than localized,
quantum dots lasers to lasers to traveling in
superconductor at individual electrons
on silicon manipulate manipulate waveguides /
the same time (Majorana particles)
wafers quantum states quantum states fiber

Example Players

IBM Intel Microsoft IonQ ColdQuanta Psi Quantum

Rigetti Honeywell Atom Computing Xanadu
Google AQT ORCA
Alibaba Computing

Researchers are working to alleviate these While many in the investment community are
challenges by ensuring calculations can happen hopeful that value can be created from these NISQ
fast enough before decoherence occurs and by machines in the near term, the ultimate goal is a
implementing error correction, which leverage other large, fault-tolerant quantum computer with error
supporting qubits to rectify the mistakes occurring correction implemented. To get there, the National
on the qubits that are being used to perform the Academy of Sciences, Engineering, and Medicine
calculations. Larger machines with more qubits are estimates that systems will need to be about five
required to build in redundancy and implement orders of magnitude larger than current machines
these codes. Additional improvements are needed with approximately two orders of magnitude
from both hardware and software to reduce these lower error rates (National Academies of Sciences,
errors and to increase the interconnection of qubits Engineering, and Medicine, 2019).
as well.
There are multiple ways to physically make gate-
Current systems are limited to tens of physical based quantum computers and many large
qubits, and no error correction is applied, meaning corporations and startups are working hard to
that there are no ideal logical qubits. Logical qubits scale various approaches. Superconducting and
are an abstraction that describes a collection of trapped ion approaches are furthest along so far;
error-corrected imperfect physical qubits that however, it is too early to speculate which specific
effectively create one qubit capable of carrying architecture will dominate, if any at all. Some
out a fault-tolerant qubit operation. There is wide believe that multiple approaches will coexist as
variation in qubit quality, which is why many refer to specialty solutions because the associated tradeoffs
these machines as belonging to the NISQ Era – i.e., may make different types of quantum computers
Noisy Intermediate-Scale Quantum (Preskill, 2018). better suited to solve specific types of problems.

31
We’ve profiled some of the most prominent approaches for building quantum computers below, but any
approach should generally satisfy the guidelines defined by David DiVincenzo in 1996 during his time at
IBM, commonly referred to as the DiVincenzo criteria, which require (National Academies of Sciences,
Engineering, and Medicine, 2019):

• Well-characterized two-level systems that can be employed as qubits

• An ability to prepare the qubits into a simple |000…> state

• Decoherence times long enough to carry out the computation or error correction

• A universal set of quantum operations (gates)

• An ability to measure qubits one-by-one without disturbing the others

Physical Platforms

There are a number of physical systems that can create platforms for quantum computing. Because
superconductors and trapped ions are presently the most advanced implementations, they are discussed
first.

Superconducting

Superconducting quantum computers are based on resistance-free

current flowing in opposite directions around a superconducting
circuit containing Josephson junctions. The qubit states are
controlled using microwave and low-frequency electrical signals.
The duration of the microwave pulses determines which quantum
operations are performed.

Several large players are working on superconducting quantum

computers, including IBM, Google, Alibaba, Rigetti, and D-Wave.
The advantages of this approach are the relatively short gate
operation times compared to trapped ion systems, and the Credit: Phil Saunders Graphics/from Optics & Photonics News, October 2016.
opportunity to build on the existing knowledge and manufacturing
base from the semiconductor industry to help scale to larger
numbers of qubits.

Some challenges are that these systems require dilution refrigerators

to keep temperatures very cold and the quantum state can
collapse easily (i.e., these systems have relatively short coherence
times). Coherence times are typically compromised further when
scaling to more qubits due to variation in fabrication methods.
Furthermore, in current superconducting implementations, the set
of qubits is not fully connected (which would allow gate operations
between any pair of qubits), restricting the kinds of algorithms that
can be performed. From Science News Feature, "Scientists are close to building a quantum
computer that can beat a conventional one" by Gabriel Popkin, Illustration by
Chris Bickel/Science. Reprinted with permission from AAAS.

32
Current superconducting quantum computers do not require optics; however, some believe that fiber optics
may be needed to support larger systems. The National Academy of Sciences, Engineering, and Medicine
estimates that current dilution refrigerator technology can handle up to ~1,000 qubits; as superconducting
systems scale beyond that limit, qubits may need to be stored in separate dilution refrigerators connected
by optical fiber (National Academies of Sciences, Engineering, and Medicine, 2019). Microwave-to-optical
transducers would be required to temporarily convert the microwave qubits to “flying” photonic qubits.

Trapped Ions

The trapped-ion approach to quantum computing involves

suspending ions (ionized atoms) in electric fields in vacuum
chambers, then using lasers to cool and prepare the ions (putting
them into well-defined initial states). To perform operations,
resonant optical or microwave fields are applied to drive qubits
into a different state or to generate entanglement between
ions. After computation, states are measured by detecting state-
dependent fluorescence.

IonQ and Honeywell are the leading companies working to

commercialize ion-trap quantum computers. One advantage of
using atoms over synthetic systems is that they are fundamentally
identical and therefore not susceptible to errors from manufacturing
Credit: Phil Saunders Graphics/from Optics & Photonics News, October 2016.

defects. This approach enables high-fidelity qubit operations and

long qubit lifetimes. A big advantage of using trapped ions as
qubits is that the set of qubits with an ion chain is fully connected
(which allows gate operations between any pair of qubits), enabling
a wide range of algorithms that can be performed.

The primary challenge for trapped ion approaches is scaling to

larger systems, which will require integration of many lasers,
vacuum systems, etc. More complex architectures are needed to
scale to larger systems beyond tens of ions lined up in a single
chain. The gate operation times are also relatively long compared
to other systems.
From Science News Feature, "Scientists are close to building a quantum
computer that can beat a conventional one" by Gabriel Popkin, Illustration by
Multiple photonic components are required for ion-trap quantum Chris Bickel/Science. Reprinted with permission from AAAS.

computers, including:

• Lasers for modulating the quantum state

• Lasers for cooling the ions, performing gates, and measurement

• Collection and imaging optics to collect the photons scattered from the ions

• Photodetectors to measure the state of ions by detecting the photons they scatter (i.e., measuring the
state-dependent fluorescence)

33
Neutral Atoms

Quantum computers based on neutral atoms are for computation. One other differentiating feature
a newer technology, but the operating principle of neutral atom systems is their capacity to support
has many similarities to trapped-ion quantum multidimensional arrays of atoms, which is more
computers. Instead of using ions, neutral atoms difficult with trapped ions due to their strong
are trapped in magnetic or optical fields (arrays of interactions.
laser light beams) in vacuum chambers. Optical and
microwave pulses are again used to manipulate the ColdQuanta and Atom Computing are two startups
quantum states of the qubits. Lasers are used to working on neutral atom approaches. Their systems
cool the atoms, which reduces computational noise. benefit from using inherently identical atoms as
their building block and have long qubit lifetimes.
Atom-based approaches were developed to They have similar scaling issues to trapped ions
address some of the challenges associated with ion because many lasers and vacuum systems need to
trap approaches. Ions, due to their charge, interact be engineered into the complete system.
strongly with magnetic fields and each other, which
can be advantageous for computation, but also The general requirements for optics and photonics
makes scaling to larger systems difficult because for neutral atom systems are similar to trapped-
undesired interactions may occur, resulting in noise. ion computers; however, different wavelengths
of light will be required to match with the atom
In contrast, neutral atoms are less prone to cross / ion transition. As in the ion trap approach,
talk and noise and are easier to scale; however, it photodetectors and collection optics are needed.
is more difficult to get them to interact as required

Photons

Photonic quantum computers, sometimes referred to as

linear-optical quantum computers, encode qubits in light
traveling in planar waveguides or optical fiber. Different
groups are implementing slightly different approaches
using either single photons (so-called discrete states) or
squeezed light states (called continuous states).

One key advantage of photonic approaches is that photons

generally interact weakly with their environment and with
each other, which increases their resistance to interference.
Two-photon quantum interference or extremely high
nonlinear optical interactions are needed to get the
photons to interact for implementing multi-qubit gates. Credit: Phil Saunders Graphics/from Optics & Photonics News, October 2016.

Psi Quantum, Xanadu, and ORCA Computing are each developing their own photonic-based quantum
computing architectures. These approaches can operate at room temperature and offer the potential for
long qubit lifetimes, fast gate operation times, and high single-qubit gate fidelity. They are also easily
networked with fiber optic communications.

Unlike other systems with stationary qubits, there are unique challenges associated with scaling to larger
systems because the photons are always moving if not stored in a quantum memory that consists of a

34
collection of absorbing atoms. Many qubits need to be routed along chips or fiber, which could impact the
final system size.

As the name suggests, photonic quantum computing approaches rely heavily on optics and photonics.
Additional improvements in components, especially single-photon sources and high-speed, low loss
switches, are critical for advancing the performance of these systems. Additional component needs include
single-photon detectors, waveguides, modulators, couplers, and filters.

Color (Vacancy) Centers

A nitrogen atom placed next to a carbon atom vacancy can add an

electron to a diamond lattice to create what’s commonly referred to as a
nitrogen-vacancy (NV) center. The spin state of the electron and nearby
carbon nuclei can be controlled with light and optically detected. The
isolated spins can be very stable and therefore used as qubits. Other
color centers, such as the silicon-vacancy (Si-V) center in diamond, offer
alternatives with advantages compared to the NV center.

Credit: Phil Saunders Graphics/from Optics & Photonics News, October 2016.

We are not aware of any commercial entities that are developing quantum
computers based on NV centers, although there are some academic labs
exploring this approach. NV centers are also being explored for quantum
memories. One advantage is that these systems could theoretically operate
at room temperature; so far, however, establishing entanglement and
reliable production of identical vacancies has been challenging. Like other
systems, NV-center-based quantum computers would require lasers and
photon detectors if commercialized.
From Science News Feature, "Scientists are close to building a quantum
computer that can beat a conventional one" by Gabriel Popkin, Illustration by
Chris Bickel/Science. Reprinted with permission from AAAS.

Silicon Spins

Silicon-spin quantum computers manipulate the spin state of

electrons confined in fabricated quantum dots on silicon wafers.

Intel is the largest player working on this approach in collaboration

with some universities. If developed successfully, silicon
spin approaches would benefit from existing semiconductor
manufacturing expertise that could allow for rapid scaling. Other
potential advantages include fast gate operation times and small
size; however, like superconducting quantum computers, cryogenic
temperatures are required to protect the quantum state.

Challenges include poor qubit uniformity and background disorder,

which is currently compensated for through careful tuning of gate From Science News Feature, "Scientists are close to building a quantum

voltages. The spin states are also highly sensitive to magnetic noise computer that can beat a conventional one" by Gabriel Popkin, Illustration by
Chris Bickel/Science. Reprinted with permission from AAAS.

35
from surrounding atoms on silicon chips; noise needs to be reduced to improve gate fidelity. Furthermore,
development of the appropriate geometry to scale to 2D arrays is needed. Aside from the optics required
for networking, silicon spin approaches are unlikely to require optics or photonics components.

Topological Qubits

Topological quantum computers maintain quantum states

in large clouds of electrons rather than localized, individual
electrons. Quasiparticles (e.g., Majorana particles) may be
created by channeling electrons through nanowires. Quantum
states are encoded by different ‘braiding’ paths that the
particles follow in time. The shapes of the braided paths lead
to superpositions and they are topologically protected from
collapse.

Microsoft is the largest player developing topological quantum

computers based on Majorana particles. Since the qubits are
protected by the topological symmetry implemented at the
microscopic level, these computers are less error-prone. There
is potential for very high fidelity and long qubit lifetimes; From Science News Feature, "Scientists are close to building a quantum
however, the technology has not yet been demonstrated. computer that can beat a conventional one" by Gabriel Popkin, Illustration by
Chris Bickel/Science. Reprinted with permission from AAAS.
While the technology is still very early stage, it has the potential
to scale readily once demonstrated and would ideally not
require error correction. It is unlikely that topological quantum
computers would require optics and photonics components.

36
OPTICS AND PHOTONICS COMPONENTS NEEDS
What are the opportunities for component suppliers?

Many optics and photonics components are required to enable quantum sensors, computers, and
communications networks (Awschalom, D., Berggren, K. K., Bernien, et. al., 2019). This chapter outlines the
outstanding needs required for photon sources (lasers and single-photon or entangled photon sources),
photon detectors, fiber and waveguides, modulators, interconnects (transducers and converters), memories
and repeaters, nonlinear-optical materials, and integrated photonics.

SUMMARY: OPTICS AND PHOTONICS COMPONENT REQUIREMENTS Required May Use

Single or Heterodyne and Fiber or

Single Transducers Memories
Entangled Homodyne Integrated Operating
Category Technology Lasers Photon Modulators and or
Photon Photon Photonic Wavelength
Detectors Converters Repeaters
Sources Detectors Waveguides

Non-
Atomic Clocks If networked
Telecom

Atom Non-
If networked
Interferometers Telecom
Sensing
NV Center Non-
If networked
Sensors Telecom

Non-
Quantum LiDAR
Telecom

Continuous
Telecom
Variable QKD

Discrete Variable
Comms. Telecom
QKD

Entanglement-
Telecom
Based QKD

Superconducting If networked --

Non-
Ion Trap If networked
Telecom

Non-
Neutral Atom If networked
Telecom

Photonic - Memory-
Telecom
Computing Discrete Variable dependent

Photonic - Cont. Memory-

Telecom
Variable dependent

Non-
NV Center If networked
Telecom

Silicon Spin If networked --

Topological TBD TBD --

Source: Expert interviews, Newry analysis

Photon Sources

Many types of photon sources are required to Lasers for Atom- and Ion-Based Quantum
enable quantum technologies. While the specific Computers and Sensors
performance requirements (e.g., wavelength, power,
linewidth, noise, single vs. entangled photons, etc.) Atom- and ion-based quantum computers and
vary widely by system, some of the most common sensors (e.g., ion trap computers, neutral atom
unmet needs that exist across multiple applications computers, atomic clocks, atom interferometers)
are outlined below. require lasers at wavelengths specifically tuned

37
to be resonant with their respective atom or ion • High repetition rate: High repetition rate (on the
transition. Lasers are used to prepare atoms, to order of hundreds of megahertz) pulsed lasers
initiate quantum states, and to read quantum are desirable for increasing sensor sampling
information. In some cases, these transitions are rates and accelerating gate operation speeds.
in the near-UV and blue-violet spectral frequency
ranges. Commercial lasers at these wavelengths (if • Compact and rugged: Smaller scale lasers
available at all) are often relatively less optimized will be critical for reducing the size and
than lasers developed for telecom applications at weight of cold atom sensors and clocks to
infrared wavelengths. Higher-performance and enable broad deployment of systems in the
more stable lasers in the near-UV and blue-violet field. Chip-integrated form factors would be
spectrum are desired to ensure precise qubit desirable. Lasers that are stable in a variety
preparation and manipulation. of environmental conditions – e.g., vibration,
moisture, temperatures ranging from 0 to 80°C,
EXAMPLE WAVELENGTH REQUIREMENTS
etc. – are also needed.

Ions – Yb+, Sr+, 369, 393, 397, 408, 422, 435, There are multiple laser technologies that may be
Ba+, Ca+ 455, 493 nm considered to meet these requirements, including:

767, 770, 780, 795, 852, 895 • External cavity diode lasers (ECDLs)
Atoms – Rb, Cs, K
nm
• Diode-pumped solid-state (DPSS) lasers

Specific requirements include: • Distributed Bragg reflector (DBR) or distributed

feedback (DFB) lasers
• Narrow linewidth: Narrow linewidth and low
phase noise are common requirements for • Vertical-cavity surface-emitting lasers (VCSELs)
lasers used in cold atom systems for quantum
sensing and computing. For applications such • Volume-holographic-grating (VHG) lasers
as probing and measurement, linewidths may
need to be as low as tens of kHz or even tens of Various suppliers such as M-Squared, Toptica,
Hz or less; requirements are most stringent for Photodigm, NKT Photonics, Sacher Lasertechnik,
optical-frequency atomic clocks. The lasers used Kelvin Nanotechnology, and UniKLasers have already
for atom or ion cooling have less demanding invested in developing lasers to meet the needs of
parameters, but still require relatively narrow quantum technology applications. Many quantum
linewidths on the order of one to tens of MHz. computing and sensing customers even claim that
recent developments have addressed the majority
• Stable output: Lasers must be mode-hop-free of their performance needs; however, reductions in
(i.e., must maintain a consistent frequency) cost (3-10x) and size are still desired. Ruggedization
and provide stable power output to ensure will also be required to use these systems in space.
consistent performance for precise qubit Multiple experts said that integrated, on-chip lasers
manipulation. would be the “Holy Grail,” particularly for quantum
sensors, though they acknowledged that it will be
• High power: Lasers used for cooling ions / atoms some time before such systems are developed.
or trapping atoms in optical lattices can require
power levels from hundreds of milliwatts up to Stability of the supply chain is also a concern. For
ten watts. Lasers used for other purposes (e.g., example, the development of ECDLs depends
probing) may only require tens of milliwatts or on the availability of laser diode components at
less. the required wavelengths, but the volumes from

38
quantum applications are often too low to motivate arbitrarily chosen by the user (unlike probabilistic
suppliers. Until demand increases, many system sources, which emit randomly and may not
designers may be limited to selecting atoms / ions emit any photons for the majority of trials). A
with transitions that match other commercially truly deterministic single-photon source would
relevant wavelengths (e.g., 780 nm lasers that enable significant advancements in photonic
were developed for CD drives align with the quantum computing.
transitions in Rb-based systems). While there are
similar wavelength requirements between these • Indistinguishable: Each emitted photon is high-
applications, many other performance specifications purity and indistinguishable from the others;
need to be met to make these lasers feasible for the frequency is the same and there are well-
quantum technologies. defined states in spatial, temporal, and spectral
modes.
Single-Photon and Entangled-Photon Sources
• Low probability of multi-photon emission: The
Single-photon sources are critical components for probability of emitting a discrete single photon
discrete-variable quantum key distribution and would ideally be 100%, with 0% probability of
some types of photonic quantum computing. Some multi-photon emission.
of the approaches used to create single photons
may also be used to create pairs of entangled • High repetition rate / brightness: Emission rates
photons, which are needed for quantum LiDAR and are very fast – i.e., many single photons can be
entanglement-based quantum key distribution, as produced per unit time.
well as emerging techniques in quantum-enhanced
microscopy and spectroscopy. • Ability to generate entangled pairs: Some
applications (e.g., entanglement-based QKD)
An ideal single-photon source would have the require the production of entangled photons.
following properties:
• Integrated: If possible, on-chip integration of
• Deterministic: Consistently emit a photon in single photon sources would be highly desirable
a controlled manner “on demand” at a time for integrated photonics.

SINGLE-PHOTON SOURCE TYPES

Deterministic Probabilistic

Multiple approaches exist but all rely on a Generate heralded pair of single photons, and the detection
similar principle: of one “heralding” photon indicates the existence of another
• An external control puts the system into photon. Usually involves laser excitation of nonlinear optical
an excited state materials, but there are two approaches:
• A single photon is emitted upon Parametric Down-Conversion Four-Wave Mixing
relaxation to the lower energy state Convert one photon of higher energy Two pump photons convert into
into a pair of photons two photons in the presence of
a third-order nonlinear material
1. Excitation 2. Emission Signal
Photon Heralding
Detector
Pump Pump Idler Photon
External Photon Idler
Control Photon Pump Signal Photon
Nonlinear
χ(2) Crystal
Source: Newry

39
 COMPARISON OF HIGHEST PERFORMANCE LEVELS ACHIEVED BY DIFFERENT SINGLE-PHOTON
SOURCE TYPES*
Spontaneous
Four-Wave Atoms and Quantum “Ideal” Single-
Down NV Centers
Mixing Ions Dots Photon Source
Conversion
Entanglement
0.9959 0.997 0.93 0.978 --- Approx. 1
fidelity
Probabilistic / Probabilistic,
Probabilistic Probabilistic Deterministic Deterministic Deterministic
deterministic Deterministic
Varies – many
Emission range 600-1700 nm 600-1550 nm Transition lines IR, telecom 600-800 nm apps require
telecom
Brightness 2.01 MHz 855 kHz 55 kHz 28.3 MHz 850 kHz High
Best g (Purity) 0.004 0.007 0.0003 0.000075 0.07 Approx. 0

HOM visibility
0.99 0.97 0.93 0.9956 0.66 Approx. 1
(Indistinguishability)

Room temp., Room temp.,

Operating temp. 273-473 K Room temp. 300-500 K Room temp.
mK (in cavity) cryogenic
Efficiency max.
0.84 0.26 0.88 0.97 0.35 Approx. 1
(Deterministic)
Metrology,
Likely quantum information, Integrated Comms, Comms,
Foundations --
apps comms, photonics foundations networks
foundations

*Note: values represent the highest performance level achieved in a single demonstration – i.e., a column is not representative of a singular system and in reality, tradeoffs
exist when optimizing versus the various requirements.
Source: OPN September 2019 Issue: Single Photon Sources, U. Sinha, S. Sahoo, A. Singh, K. Joarder, R. Chatterjee, S. Chakraborti
Links to original sources can be found at: http://www.rri.res.in/quic/resources/opn2019

There are five main technological approaches that • Four-Wave Mixing: Four-wave mixing in a glassy
may be used to generate single and / or entangled (third-order) nonlinear optical material is a
photons; however, each has its tradeoffs: waveguide-based approach that is often used
in integrated photonic platforms. It offers high
• Spontaneous Parametric Down-Conversion entanglement fidelity but is not deterministic.
(SPDC): SPDC in a crystalline (second-order)
nonlinear optical material is one of the most • Atoms and Ions: Atom and ion sources can be
popular approaches for generating single used to create deterministic photon sources with
photons. It generates photons at a fast repetition highly indistinguishable output, narrow emission
rate (~2 MHz), often defined as brightness, and linewidths, and high purity. Because atoms emit
provides high entanglement fidelity (>99.5%); isotropically, very high-quality optical cavities
however, it is not deterministic and there is are needed to make these sources truly or nearly
some probability that multi-photon emission deterministic. This approach is not considered
can occur (Sinha, et al., 2019). Coupling into the optimal choice for many commercial
optical fiber can also be a challenge. applications due to the complexity of the atom/

40
 ion trap system and the high-quality cavity that Discrete-variable photonic quantum computing
is required but may be used in experimental approaches (e.g., PsiQuantum) require deterministic
research. single-photon sources that have high single-photon
purity, high photon indistinguishability, and high
• Quantum Dots: Quantum dot-based single- brightness. The source should also be compatible
photon sources offer the highest brightness with integration into CMOS fabs. Other photonic
(single photon emission rates of 28+ MHz) and quantum computing approaches (e.g., Xanadu)
have the potential to be deterministic (Sinha, may alternatively use squeezed light sources based
et al., 2019). They also have relatively high on nonlinear optics.
entanglement fidelities (97.8%) and purity,
but may require cryogenic temperatures to Entangled photon sources are required for
operate optimally (Sinha, et al., 2019). Many entanglement-based QKD protocols and specialized
believe this is a promising approach for applications in metrology and imaging.
deterministic generation of indistinguishable
single photons, although generating identical
quantum dots remains a challenge and most
commercial sources are still probabilistic. If
two quantum dots have significant structural
differences, the photons emitted from each
will be distinguishable. Coupling into optical
fiber can also be a challenge. Multiple startups
(e.g., Quandela, Sparrow Quantum) are actively
commercializing the technology.

• Nitrogen-Vacancy (NV) Centers: Color centers

in diamond, most commonly known as NV
centers, are a newer approach to single-photon
generation. They are in principle deterministic
sources and the emission rate increases with
temperature, which makes them potential
candidates for room-temperature sources.
Consistent manufacturing of NV centers remains
a barrier.

Given the tradeoffs described above, different ap-

proaches may be preferred for different applica-
tions based on the varying requirements.

For example, discrete-variable QKD protocols

require bright and high-purity sources to increase
key exchange rates and maintain security,
respectively. Room temperature operation is also
preferred to minimize costs. The preferred emission
wavelength is ~1550 nm for fiber-based systems
and other spectral ranges may be considered for
free-space, satellite-based networks to minimize
scattering and maximize transmission.

41
Photon Detectors

Photon detectors are also widely used for many quantum technologies, ranging from measurement of state-
dependent fluorescence in trapped-ion quantum computers to detection of single photons transmitted in
QKD and photonic quantum computers. Two general photodetector categories are outlined in more detail
below: single photon detectors and heterodyne / homodyne detectors.

Single-Photon Detectors

Single-photon detectors are required for discrete- high efficiency, low dead time, low timing jitter,
variable QKD and photonic quantum computing, and low dark count rates; however, they require
some types of quantum imaging, and quantum cryogenic cooling, which can be cost-prohibitive
LiDAR. In addition to quantum technologies, other for some applications. Avalanche photodiodes
applications (e.g., ultra-low light sensing and (APDs) operating in Geiger-mode are a more
surveillance, medical imaging, astronomy) may mature technology that is used in applications such
benefit from their development. as QKD and does not require cryogenic operation.
Lastly, photomultiplier tubes (PMTs) are another
An ideal single-photon detector would have the very mature technology, but their use in quantum
following performance specifications: applications is relatively limited.

• High detection efficiency: There is high (close to SSPDs, APDs, and PMTs are not intrinsically photon-
100%) probability that a photon incident upon number resolving; however, superconducting
the detector is successfully detected – i.e., no transition edge sensors (TES) and CMOS image
false negatives. sensors that provide this capability are in
development. Bolometric TES require cryogenic
• Low dead time: The time after a photon- temperatures but can offer photon-number-
detection event during which the detector resolving capability, high efficiency, and low dark
cannot detect a photon is negligible (near-zero). count rates. CMOS image sensors could operate at
room temperature and have low dead times, high
• Low timing jitter: There is minimal or no variation efficiencies, and low dark count rates; however,
from event-to-event in the delay between the the long readout time may be limiting, and current
input/output signals. silicon-based imagers are not compatible with
telecom wavelengths.
• Low dark count rate: There would ideally be
no detector output pulses in the absence of Geiger-mode avalanche single-photon detectors
incident photons – i.e., no false positives. are commercially available from players such
as ID Quantique, Aurea Technology, Micro
• Photon number resolving: Some applications Photon Devices, and PicoQuant. Wooriro is a key
require a detector that can distinguish the supplier of InGaAs/InP avalanche photodiodes.
number of photons in an incident pulse rather Superconducting single-photon detectors, a newer
than just providing a “0” or “≥1” reading. technology, are also already being supplied by
several companies, including Quantum Opus, Single
There are several technological approaches to Quantum, and Photon Spot. Research is ongoing to
meet these requirements. Superconducting single- further improve performance, integrate on-chip,
photon detectors (SSPDs) are a newer technology and reduce size / cost.
that is gaining commercial traction. SSPDs offer

42
OVERVIEW: SINGLE-PHOTON DETECTOR TECHNOLOGIES
Type Status Cost Advantages Drawbacks
• Low detection efficiency
Photomultiplier Very • Room temperature • High timing jitter
$
Tubes (PMTs) Mature • Mature, low-cost technology • High dark count rate
• Not photon number resolving
• Incompatible with IR (Si is
transparent); InGaAs may be used,
• Room temperature but lower efficiency and higher dark
• Lower size, weight, and power count rates
vs. SSPDs • Not photon number resolving,
Avalanche
Photodiodes Mature $ • Moderate timing jitter (high unless multiplexed
(APDs) time resolution) • High dark count rates and/or after-
• High count rates and room pulses, especially in NIR; lower dark
temperature operation can be count rates possible with coolers
achieved in fast gating mode • Lower efficiency vs. SSPDs
• Longer dead times
• Low timing jitter
• Requires cryogenic cooling (<4K)
• High detection efficiency
• Not photon number resolving,
Superconducting • Low dead time
Small unless parallel configuration
Single Photon $$$
Volume • Low dark count rate • Susceptible to picking up
Detectors (SSPDs)
• No after-pulsing background thermal radiation from
input fiber
• High photon detection rate
• Requires ultra-low-temperature
• High detection efficiency
Transition Edge Early operation (tens of mK)
$$$ • Photon number resolving
Sensors (TES) Stage • High timing jitter
• Low dark count rates
• Longer dead times
• Room temperature
• Photon number resolving • Incompatible with IR (Si is
CMOS Image Early transparent), lower efficiency in
$ • Low dead time blue/UV
Sensors Stage
• High detection efficiency • Long readout time (microseconds)
• Low dark count rate

Heterodyne and Homodyne Photon Detectors

Other applications that do not use single photons, such as continuous-variable QKD and atom-based
computing and sensing systems may use coherent heterodyne or homodyne detectors. In general, our
interviews revealed that there are relatively fewer unsolved problems / pain points associated with current
heterodyne and homodyne detectors compared to single-photon detectors. This could be because some
of these detectors are already used in non-quantum (classical) communications today. Lower noise and
higher bandwidth detectors would be a valuable advancement.

43
 COMPARISON OF COMMON COMMERCIALLY AVAILABLE INFRARED SINGLE-PHOTON DETECTORS

Avalanche Photodiode Superconducting

“Ideal” SPD
InGaAs / InP, Geiger-Mode Nanowire SPD
Example Product ID Quantique ID230 Single Quantum Eos --
Detector Efficiency at 1550 nm 10 - 25% ≥ 85% 100%
Reset Time 2 - 25 μs ≤ 30 ns Approx. 0+
Jitter Time 150 ps ≤ 25 - 50 ps 0.0 ps
Dark Count Rate <50 – 200 Hz ≤ 300 Hz 0.0 Hz
Non-Cryogenic (183 -
Operating Temperature Cryogenic (~2.5K) Room Temp. (~293K)
223K)
Relative Cost Low High --
Intrinsically Photon Number
No No Yes
Resolving
Sources: Data sheets from Single Quantum, IDQuantique

Components for Transmission and Light Manipulation

Many additional optics and photonics components are required to manipulate and transmit the light
between the photon source and detector. Specific needs by component type are profiled in more detail
below. One common requirement across most components and applications is the need for ultra-low-loss
components; given that many quantum systems are based on manipulating single photons, truly every
photon counts.

Fiber / Waveguides

Lower-loss fiber and waveguides would improve losses are currently on the order of 0.25 dB/cm
the performance of quantum communications in Si and 0.1 dB/cm in SiN, but even lower-loss
networks, photonic quantum computers, and waveguides (down to 0.01 dB/cm if possible) are
chip-based sensors. Fiber may also be used to desired. Multiple approaches are being trialed to
network between dilution refrigerators and ion further reduce loss, including new materials and
traps in superconducting and trapped ion quantum development of optical coatings and improved
computers, respectively. Quantum communications manufacturing processes (e.g., reducing side wall
and photonic quantum computing approaches width variance in integrated photonics).
may operate at telecom (infrared) wavelengths,
while atom- and ion-based sensors may require a In addition to being lower loss, fiber or waveguides
different spectral frequency range (commonly blue for systems that use photons as the qubit (e.g.,
or UV) and therefore potentially different materials. photonic quantum computers, QKD) must be
polarization-maintaining. The industry also
In existing ultra-low-loss commercial fiber, losses expressed interest in commercial availability of low-
are ~0.17 dB/km, but quantum communications loss hollow fibers filled with reactive alkali-vapor
applications would benefit from additional atoms for atom-based quantum memories.
reduction. For integrated photonic approaches,

44
Modulators and Switches reflection, etc. Spatial light modulators may also be
used for high-dimensional (multi-state) QKD.
Modulators are another enabling technology used
across computing, communications, and sensing Many existing modulators suppliers serve traditional
platforms for switching and to modulate phase, markets (e.g., telecom) and university labs and
amplitude, polarization, frequency, etc. other companies are actively working to advance
performance further for quantum applications. For
Low-loss, high-speed electro-optic modulators are example, QUBIG GmbH was founded in 2008 by two
needed to ensure fast switching and routing for quantum optics researchers and is now supplying
photonic quantum computing, which relies on large, electro optic modulators to multiple quantum
multiplexed optical circuits and rapid feedforward. projects in Europe. Hyperlight, a startup spun out
While high-speed modulators are available today, of Harvard, is commercializing integrated electro
losses are typically too high (~2-3 dB) for quantum optic modulators based on thin-film lithium niobate.
applications, which cannot use amplifiers and Other example suppliers include EO Space, Gooch
therefore require losses under a few tenths of & Housego, and Harris Corporation.
dB. One expert described the ideal, even “game
changing” modulator as one that is waveguide- or Interconnects: Transducers and Converters
fiber-based, has a speed of a few GHz, and has very
high transmission (99%) at telecom wavelengths. Optical frequency converters and optical-to-
microwave transducers will be needed in the future
MODULATOR PERFORMANCE TARGETS to link disparate systems in a quantum network
to create a quantum internet or local network of
Performance Metric Current Ideal quantum computers (Awschalom, D., Berggren, K.
Loss 2-3 dB <0.1-0.5 dB K., Bernien, et. al., 2019). Quantum interconnects
must transfer the quantum states between various
Speed A few GHz A few GHz physical media (e.g., atoms, photons, microwave
Source: Expert interviews
fields, semiconductor electronics) with high fidelity,
fast rates, and low loss.

Modulators are also critical components for Interconnects will be required to link both bosonic-
quantum key distribution, which requires high- to-atomic and bosonic-to-bosonic systems
bandwidth, low-voltage, and low-loss modulators. (Awschalom, D., Berggren, K. K., Bernien, et. al.,
These components need to operate at telecom 2019). Bosonic systems include optical photons,
wavelengths and are used to modulate amplitude, mm waves, microwaves, and acoustic photons.
phase, and / or polarization of light. Atomic systems include neutral cold atoms, trapped
ions, quantum dots, color centers, superconducting
Atom- and ion-based quantum computers and Josephson devices, etc.
sensors / clocks also use electro-optic phase,
amplitude, and polarization modulators for Microwave-to-optical transduction will be required
laser cooling, quantum state manipulation, and to network matter-based quantum computing
readout. Unlike photonic computing and QKD, approaches (e.g., superconducting) with fiber
these applications may require operation in blue optics because photons are the most practical
/ UV wavelengths. I/Q (quadrature) modulators non-stationary qubits that can be transmitted over
are also desired at these wavelengths. Neutral long distances. This networking between quantum
atom quantum computers specifically require computing or sensing nodes is likely to occur in the
higher-bandwidth (up to tens of GHz) spatial light quantum internet.
modulators with greater power handling capacity
(up to 1 W) for spatial filtering, acousto-optical

45
Optical fiber may also be used to link multiple Meeting the efficiency targets for microwave-to-
dilution refrigerators for superconducting quantum optical transduction remains challenging for all
computing architectures as they scale to higher- implementations. While optomechanical systems
qubit systems. Although these systems operate in have the highest demonstrated efficiencies,
the microwave regime, even low-loss microwave they suffer from thermal noise and tend to have
cables experience high attenuation (>1 dB/m). low bandwidth. Other systems have seen large
By converting between microwave and telecom improvements over the last decade and may be
frequencies, networks of matter-based systems able to offer competitive efficiencies and wider
can take advantage of the dramatically lower loss bandwidths in the future.
telecom optical fiber (<0.2 db/km).
Even if non-microwave systems are being used
Requirements for microwave-to-optical transducers (e.g., trapped ion, NV-center), multi-frequency
include high fidelity, high efficiency (as close to conversion may be required if the systems operate
100% as possible, but must be above 50%), and at different wavelengths (e.g., visible-to-infrared).
bandwidth of at least 10 kHz. Transducers also must For example, NV centers, which emit at 637 nm, may
be able to operate at millikelvin temperatures and be used as network nodes or distributed sensors
have low power dissipation to be compatible with in the quantum internet; however, conversion to
the cryogenically cooled superconducting systems telecom wavelengths (e.g., 1550 nm) is needed to
(Lambert, Ruedo, Sedlmeir, & Schwefel, 2020). transmit between nodes over traditional fiber optic
cable. Quantum memories (described in more detail
A wide range of transducer implementations below) often emit at non-telecom wavelengths as
are being considered, including those that well, so frequency conversion will be required in
leverage electro-optical, magneto-optical, and some memory-based quantum repeaters, once
optomechanical effects. Systems based on developed.
absorption spectra of rare-earth ions and Rydberg
systems are also being explored.

46
The quantum state must be maintained during Photonic quantum repeaters would obviate the
the conversion process, and the converters must need for memory by creating “cluster states,” large
be high efficiency (ideally 95-100%), low loss, and groups of highly entangled photons, and then
low noise. Key research goals include reducing performing specific operations and measurements
coupling loss and increasing conversion efficiency on those qubits. Rudimentary photonic-based
while preserving the quantum properties of the quantum repeaters may be able to be implemented
converted photons. at an earlier stage.

Multiple approaches are being pursued to meet Quantum memories can be applied in quantum
these performance targets. The most common computing as well as communications. The memory
approaches rely on materials with 2nd order (e.g., stores the qubits (e.g., single photons) temporarily
lithium niobate) and 3rd order (e.g., silicon nitride) while other processing occurs and then the quantum
optical non-linearities where laser pumps are used state is recreated for subsequent processing. The
to up- or down-convert the frequency of the signal memory must hold multiple pieces of information in
photon. Bulk crystals may be used for this purpose parallel, and the information cannot be observed.
and near-100% efficiency is possible, but high-
power lasers are required, which generate noise via Quantum memories may also be combined
Raman scattering, thereby making it more difficult to with probabilistic single-photon sources (e.g.,
isolate the signal photon. Alternatively, waveguides spontaneous parametric down conversion) to
may be used, which enable the use of lower power temporarily store single photons, thereby creating
lasers; however, on- and off-chip coupling losses an effectively deterministic single photon source
are a challenge. Additional approaches include ring as long as the memory is high-efficiency. There are
resonators and electro-optical conversion, which several key parameters that must be met:
can achieve high efficiency conversion but are
limited to smaller frequency shifts. • High efficiency and fidelity: The ideal quantum
memory would have write-in/write-out
Quantum Memories and Repeaters efficiencies exceeding ~90%, meaning that
the more than 90% of the input information is
Quantum repeaters are a critical element for successfully output.
enabling quantum communications over longer
transmission distances. Due to transmission loss, • Multiplexed: Storing multiple quantum states
current fiber-based QKD installations are limited to in various modes is desired – some experts
distances on the order of a few hundred kilometers. estimate that the ability to multiplex in excess
While amplifiers are used in classical communications of hundreds or thousands of modes will be
networks to mitigate transmission loss, they cannot required.
be used in quantum networks because they would
destroy the quantum state. Accordingly, new • High storage lifetime: The qubit storage
quantum repeaters are in development. lifetimes must be long – on the order of 500
microseconds to 500 milliseconds. Long storage
There are two main classes of quantum repeaters: lifetime is particularly critical for quantum
memory-based and photonic-based. Memory- communications, which has requirements on
based repeaters use a quantum memory to the millisecond or second time scale. Photonic
temporarily store quantum states while the system quantum computers can tolerate shorter storage
completes some other processing in the network, times, but other features, such as multi-mode
similar to memories used on classical computers. capacity, become more important.
Once this processing is complete the repeater can
pull the quantum state out of its memory and use
that qubit in additional processing.

47
• Compatible with networks: Most memories will that have been stored a re-emitted] due to their
need to interface with telecom networks; the large resonant optical depth, have long storage
memory must couple with the infrastructure lifetimes, and operate at convenient near-IR
effectively. transition wavelengths. They can also multiplex
over 200 modes. While systems can achieve
There are multiple physical approaches to creating good storage lifetimes at room temperature,
quantum memories. Most are based on resonant cooling vapors to mK temperatures or confining
(first order) and Raman (second order) interactions them in magneto-optical traps (MOTs) can
in atoms that may be in crystalline and amorphous dramatically improve lifetimes, in some cases
solids, molecular gases, metamaterials, etc. Example to tens of seconds. Scaling and efficient in- and
technologies that are in development include: out-coupling remains a challenge for atomic gas
memories, so integration into optical cavities or
• Rare-Earth Ion-Doped Solids: Rare-earth ion- hollow core fiber are also being explored.
doped solids use crystals (e.g., yttrium or
lanthanum) or optical fiber doped with rare- • Optical Fiber / Waveguide Memories: Since
earth ions (e.g., Nd, Pr, Eu, and Er). These photons travel through fiber at a set speed,
systems can demonstrate excellent storage long lengths of fiber or loops of fiber with a
lifetimes at cryogenic temperatures – in some switch can be used to effectively delay photons
cases, on the order of hours – and efficiencies while other computational tasks are completed.
of 70%. They are also compatible with spatial, Fiber-based memories are compatible with
temporal, and spectral multiplexing. These telecom networks and are multiplexable. The
memories can be used for systems operating key challenge is transmission loss because long
at telecom wavelengths, as well as microwave fiber distances are required to achieve the
frequencies (e.g., for superconducting qubit necessary storage times. Even low-loss fiber will
systems). only be able to hold a photon for a few hundred
microseconds before loss drives the system
• Nitrogen-Vacancy Centers: Nitrogen-vacancy efficiency below 50% (3dB loss). Switch-based
centers, often in diamond crystals, can be used as approaches face the same issue.
a quantum memory with a long storage lifetime.
Other advantages include the ability to optically While there are many research groups actively
prepare and read out states and the potential working on developing quantum memories and
for implantation in optical micro / nanophotonic repeaters, additional development will be required
devices. In addition to storing optical photon before a product can be commercialized. Similar
states, nitrogen-vacancy memories are also able to quantum computing, no single approach
to couple to and store microwave photons from has emerged as the most promising candidate
superconducting qubit systems. technology yet. Some form of error correction
will likely be required as well. An NSF workshop
• Raman Scattering in Diamond: The high Raman concluded that there are likely to be three
gain of diamond and its broad transmission generations of quantum repeaters that will bring
window could make it useful as an optical incremental performance improvements over the
memory; however, the short storage lifetime next 3-10 years (Awschalom, D., Berggren, K. K.,
(picoseconds to a few nanosceconds) would be Bernien, et. al., 2019).
insufficient for most applications.
Nonlinear Optics
• Atomic Gases: Clouds of atomic gases, typically
alkali elements like Cs and Rb, can be used to Many of the technologies discussed above will
store quantum states. These systems can achieve require advanced development of nonlinear optical
high efficiency (~80%) [the majority of photons materials and devices (Awschalom, D., Berggren,

48
K. K., Bernien, et. al., 2019). Spontaneous down
conversion and four-wave mixing as single-photon
sources need to be developed further to provide
high efficiency, wavelength tunability, linewidth
control, etc. Optical squeezed-state generators
are fairly well advanced (and are deployed
already in the laser interferometer gravitational-
wave observatory [LIGO] systems) but need to be
developed on-chip for size and cost considerations.
Optical frequency converters (including optical-to-
microwave transducers) rely on nonlinear-optical
processes and need to be further optimized and
commercialized for specific applications.

Other

The components described above represent

the needs that were most commonly expressed
throughout our interviews. Other system-specific
optics and photonics needs were also identified,
including:

• On-chip, high-extinction ratio (up to 80-100

dB), low-loss optical filters

• High-collection efficiency optics to couple light

into detectors for atom- and ion-based sensors,
clocks, or computers

• Lower-loss, smaller isolators (including at non-

telecom wavelengths for atom-based sensors)

• Small or chip-scale optical frequency combs

(low-power fiber combs)

49
Integrated Photonics: Linear and Nonlinear

Integrated photonics – linear and nonlinear optical There are many different material platforms (e.g.,
– is commonly cited as another set of enabling silica, silicon, silicon nitride, silicon carbide, lithium
technologies that would provide many essential niobate, indium phosphide, diamond, etc.) being
benefits: developed for linear and nonlinear integrated
photonics and each brings its own advantages and
• Devices with lower size, weight, power, and cost tradeoffs. Silicon-based platforms are preferred by
(SWAP-C) the industry where applicable because they can
leverage the existing installed manufacturing and
• Greater scalability knowledge base; however, different materials may
be needed to minimize loss and to integrate certain
• Improved stability of optical elements components. Polarization-maintaining materials and
components define another common requirement
• Greater ability to interface with CMOS in some applications.
electronics
Transmission loss varies by wavelength in different
• Potential for enhanced light-matter interaction materials; while QKD and photonic quantum
computing chips will likely use telecom wavelengths
Lower SWAP-C is particularly important for mobile (1550 nm), many atom-based systems require
applications – e.g., field-deployable quantum different materials optimized for the blue- or UV-
sensors and quantum key distribution transceivers. spectrum. Silicon’s bandgap is 1.1 micron, making
In photonic quantum computers, maintaining it unsuitable for chips operating in the UV-visible
amplitude and phase stability are key advantages spectrum, which require a wide bandgap for optical
gained by integrating many components on one transparency. Alternate materials such as III-V Nitride
chip to reduce the number of parts. In addition to semiconductors (which are more transparent) and
these applications, integrated photonics platforms / or other crystalline materials (which have lower
are also being developed for non-quantum optical absorption and autofluorescence) may be
markets such as classical communications (e.g., preferred, but these materials are relatively less
transceivers), biophotonics (e.g., lab-on-a-chip, developed.
optical biosensors), LiDAR, and more.

COMPARISON OF COMMON PHOTONIC INTEGRATED CIRCUIT MATERIAL PLATFORMS

Refractive
Component Propagation Thermo-optic Compatible
Material Index
Integration Loss Coefficient with CMOS
Contrast
Filters, modulators, Relatively
Silicon High High Yes
switches High
Filters, modulators,
Silica on Silicon Low Low Low Yes
switches, splitters
Lasers, amplifiers,
III-V Relatively
modulators, Low High No
Semiconductors High
detectors
Ideal Material Many or all Low High Depends Yes

50
 EXAMPLE MATERIAL OPTIONS FOR ATOM- AND ION-BASED SYSTEMS

1100 nm
Wavelength Range

The bandgap of the material must be lower than the Range of

operating wavelength of the application*
(nm)

Operating
365 nm Wavelengths
200 nm 230 nm 250 nm
140 nm

SiO2 AlN Diamond Si3N4 GaN Si Atom- and

Ion-Based
Systems

Bandgap Wavelength by Material Application Requirement

*Applications require materials that are transparent at the operating wavelength. Above the bandgap wavelength, the materials listed start to become
transparent and could therefore be considered; however, loss must also be considered

On-chip integration of different components, such as lasers and photodetectors, remains a challenge for
different material platforms. For example, on-chip integration of pump lasers has been demonstrated in
indium phosphide, but not CMOS-compatible platforms.

Most experts argue that heterogeneous integration of multiple material platforms will be required to
integrate all the necessary components on-chip. While many approaches (e.g., flip-chip bonding, full wafer
bonding, transfer printing, electron-beam lithography) are being developed to accomplish this goal, other
experts argue that the manufacturing complexity (e.g., optimizing alignment, adhesion, thermal budget
management) is a challenge that needs to be addressed to do this at-scale and realize this functionality.

Different materials may be advantageous for different device types. For example, integrated photonic chips
for quantum frequency conversion would optimally use a material that is low loss, has a wide transparency
window, and has a fast switching mechanism.

DEMONSTRATED INTEGRATION OF DIFFERENT QUANTUM PHOTONIC COMPONENTS ON CHIP BY

MATERIAL PLATFORM*
Single- Correlated
Passive Active Fiber
Pump Photon Photon Pair Waveguide Detector
Elements Elements Coupler
Source Source
Silicon-on-Insulator ✓ ✓ ✓ ✓ ✓ ✓
Silica ✓ ✓ ✓ ✓ ✓ ✓
Silicon Nitride ✓ ✓ ✓ ✓ ✓ ✓
Indium Phosphide ✓ ✓ ✓ ✓
Gallium Arsenide ✓ ✓
Diamond / DOI ✓ ✓ ✓ ✓ ✓
Lithium Niobate ✓ ✓ ✓ ✓ ✓ ✓
*Note: Simplified visual – different approaches and different performance levels were achieved in different materials
Source: Reprinted with permission from © The Optical Society. Opt. Mater. Express 7, 111 (2017)

51
 QUANTUM FREQUENCY CONVERSION: COMPARISON OF INTEGRATED PHOTONICS PLATFORMS
Transparency Nonlinear Demonstrated Single Quantum Tuning / Switching
Platform
Window Coefficient Optical Loss Emitter Integration Mechanism
Silica >140 nm Weak chi(3) Ultra-low Not native Thermo-optic
SiN >350 nm Moderate chi(3) Low Not native Thermo-optic / MEMS
Silicon-on- Nascent (Se Thermo-optic / free-
>1000 nm Strong chi(3) Medium
Insulator defects) carrier / MEMS
LiNbO3-on- Strong chi(2); Rare earth Electro-optic /
>300 nm Low
Insulator Moderate chi(3) incorporation piezoelectric
Moderate chi(2); Electro-optic /
AlN >200 nm Low Not native
Moderate chi(3) piezoelectric
GaAs-on- Strong chi(2);
>750 nm Medium InAs quantum dots Piezoelectric
Insulator Strong chi(3)
Via electron beam Electro-optic /
SiC-on- Moderate chi(2);
>400 nm (4H) High irradiation, ion piezoelectric / DC
Insulator Moderate chi(3)
implantation Stark-shift
Diamond-on- Ion implantation,
>250 nm Moderate chi(3) Medium Strain
Insulator CVD
Si or GaAs on
Rare-Earth- Rare earth DC Stark-shift /
>300 nm -- Low
Doped incorporation Zeeman shift
Oxides
Strong chi(2) =
Ideal Wide electro-optic effect Ultra-low Integration possible Electro-optic is fastest
for fast switching
Source: Awschalom, D., Berggren, K. K., Berinien, et. al. (2019). Development of Quantum InterConnects for Next-Generation Information Technologies. arXiv preprint arXiv:1912.06642.

Chip-to-Fiber and Chip-to-Chip Couplers couplers couple light in and out of the lateral side
of the waveguide and light propagates in-plane. In
Chip-to-fiber couplers and chip-to-chip couplers / contrast, grating couplers use diffraction gratings,
interposers are also critical, and often overlooked, as their name suggests, to couple light out-of-plane
components for integrated photonic platforms. (i.e., the light is incident from the top surface of the
Significant improvements are needed to reduce silicon chip).
coupling loss to make these components viable for
quantum technologies. While progress has been Both approaches are in development and have their
made to reduce waveguide losses (e.g., <0.1 dB/ own advantages and tradeoffs. Higher coupling
cm loss achieved in SiN), coupling losses on the efficiencies and lower polarization dependence
order of 3-5 or more dB still need to be reduced. are possible with edge couplers, whereas grating
Matching the size of the device mode to the size of couplers are more compatible with high-volume
the fiber mode is critical for achieving high coupling fabrication. Evanescent coupling schemes are also
efficiency. possible, where a small size optical waveguide
is coupled to a large waveguide through an
There are two general categories of chip-to-fiber intermediate mode size waveguide.
couplers: edge couplers and grating couplers. Edge

52
COMPARISON OF FIBER-TO-SILICON PHOTONIC INTEGRATED CHIP COUPLER TYPES
Coupler Type Advantages Challenges

Edge (In-Plane) • Relatively high coupling efficiency Requires relatively complex

• Broad coupling bandwidth fabrication and assembly procedures
(wavelength) (not always compatible with wafer-
scale testing)
• Low polarization dependence

Grating (Out-of-Plane) • More compatible with high-volume • Relatively lower efficiency

fabrication – no cleaving, facet- • Relatively narrower bandwidth
polishing, etc.
• Polarization-dependent; requires
• Allows for access to any part of more complex grating if multiple
optical circuit (important for wafer- states of polarization
level testing)
• Higher fiber-positioning tolerances

Source: Newry

53
OIDA SURVEY: CHARACTERIZING WHAT’S NEEDED
While researching and writing this report, OIDA conducted a survey with academic
and industry experts in the fields of quantum sensing, communications, and
computing. This data, gathered from 60 individuals who completed the survey,
was used to develop many of the conclusions in the roadmap.
Part of the survey was focused on characterizing the criticality and type of
improvements that are needed across different components.

IMPORTANCE PERFORMANCE COST

% of FOR SYSTEM IMPROVEMENTS REDUCTION
respondents: PERFORMANCE NEEDED NEEDED

Lasers n = 37

Single- or entangled- n = 28
photon sources
Single-photon detector n = 29

Heterodyne or n = 19
homodyne detectors
Optical fiber n = 36

Modulators n = 33

Microwave-to- n = 10
optical transducers
Frequency converters n = 21

Memories or repeaters n = 11

Integrated photonics n = 23

Atom sources n = 12

Control electronics n = 38

Cryogenic systems n = 25

Vacancy centers n=6

Vacuum technologies n = 28

Minimal Low Moderate High Significant

Note: Respondents were only asked to evaluate components that they currently specify in the
systems they work on
Source: OIDA/Newry Quantum Technology Survey – February 2020

54
TRACKING PROGRESS
What “beacon” milestones or metrics will indicate progress?

While early products are or are being commercialized today (e.g., atomic clocks, QKD networks, NISQ
quantum computers), overall the quantum technology market is nascent. We are many years away from
a fault-tolerant, universal quantum computer or a quantum internet. To help the investment community
evaluate progress toward these ultimate breakthroughs, we’ve outlined key technical milestones and
metrics to monitor (where possible and appropriate) across each of the technology categories below.

Quantum Sensing in length if thermoelectrically cooled semiconductor

detectors are used, or up to ~400 kilometers if
Given the breadth of technological approaches and cryogenically cooled superconducting detectors
applications for quantum sensors, it is difficult to are used (Quantum Flagship, 2019). Increasing
isolate metrics that are relevant across all segments. transmission distance will be critical to expand the
The utility of quantum sensors is dependent on the utility of QKD to more applications.
performance improvements that they are able to
deliver relative to the specific incumbent technology Additionally, current QKD rates are limited to
used in any given application. Furthermore, several megabits per second. A 2019 NSF workshop
different applications will be optimized for different set a three-year target of quantum-secured
tradeoffs, such as sensor size / weight, resolution, communication rates >1 MB/sec over 100 km
and precision; therefore, no absolute metrics are distance (Awschalom, D., Berggren, K. K., Bernien,
appropriate. Key market drivers include usability, et. al., 2019). Different rates will be required
portability, environmental stability, performance vs. depending on the quantum communications
incumbent approaches, and price. application. QKD can tolerate relatively slower
key exchange rates on the order of a few keys per
Despite this fragmentation at the level of specific second, whereas distributed quantum computing
performance metrics, wafer-scale integration of applications would require much faster data rates.
these systems is a more general technological In addition to rate and distance, cost is another
milestone that would enable progress across metric to monitor, as current QKD system costs are
multiple applications. Lower SWAP-C devices are prohibitively high for many applications.
needed across multiple sensing categories, and most
experts agree that integration of these systems onto The commercialization of a quantum repeater
photonic chips is a critical path to doing so. While would be an important technological milestone
some integration is possible today, more on-chip and may enable revolutionary progress in quantum
functionality (e.g., sources, modulators, switches) communications including remote or distributed
is needed. For example, integrated lasers with the quantum computing. As described in the previous
appropriate performance levels (e.g., mode hop- chapter, various physical approaches are being
free, narrow linewidth, temperature stable) would pursued to develop a quantum repeater; however,
be a significant advancement. Some experts even it will be several years before a quantum repeater is
described this as the “Holy Grail.” commercialized.

Satellite-based QKD development is also

Quantum Communications ongoing as an alternative approach to increasing
transmission distance, but data rates need to
Transmission distance and key / data rates are the improve further. Incremental improvements may
primary metrics to evaluate progress for quantum come from improved components such as lower-
communications. Current technology is limited to loss fiber optics and lower-noise, higher-efficiency,
fiber links on the order of a few hundred kilometers faster detectors.

55
Quantum Computing

Multiple metrics can be considered to evaluate

the progress and performance levels of quantum • Gate operation time: Gate operation time is
computers and not all are mutually exclusive. Some the clock cycle time required to manipulate the
of the most common metrics include: physical qubits for the operation.

• Qubit count (physical and logical): The number • Qubit lifetime (coherence time): Qubit lifetime
of physical qubits in a system – e.g., the number is the time that information can be stored and
of ions in a trapped-ion quantum computer. processed in the qubit, and therefore used for
Logical qubits are an abstraction that describes computing, before it decoheres.
a collection of error-corrected imperfect
physical qubits that can effectively represent a • Inter-qubit connectivity: Qubit connectivity
single ideal mathematical qubit with which to defines the number of connections between
participate in a fault-tolerant qubit operation. qubits in the array. The higher the connectivity,
Qubit count is a measure of scaling and is an the easier it is to implement quantum algorithms.
important indicator of achievable operational
complexity. • Circuit depth: Circuit depth is the number of
operations that can be performed before an
• Gate fidelity (1- and 2-qubit): Gate fidelity uncorrectable error is observed.
is the accuracy of a gate operation and is a
measurement of the qubit quality and the • Quantum volume: Quantum volume is a figure of
quantum control used to carry out the gate merit developed by IBM that approximates how
operation. Gate fidelities need to reach a powerful a quantum computer is by accounting
minimum threshold to implement quantum for qubit count, gate and measurement errors,
error correction. device cross talk, device connectivity, and
circuit compiler efficiency. IBM has doubled
• Gate set: Different systems have different gate the quantum volume of its quantum computers
sets, meaning they are able to implement every year since 2017.
different elementary 1- and 2-qubit operations.

THE MOORE’S LAW(S) OF QUANTUM COMPUTING?

NEVEN’S LAW ROSE’S LAW

At Google’s Quantum Spring Symposium in
May 2019, Hartmut Neven, Engineering
Director, stated that quantum computers are
gaining computational power at a “doubly
exponential” rate, surpassing Moore’s law

Year Moore’s Law Neven’s Law

1
Year 1 21 = 2 22 = 22 = 4
2
Year 2 22 = 4 22 = 24 = 16 Source: Media courtesy of D-Wave Systems

3 Rose’s law is attributed to Geordie Rose (former CTO,

Year 3 23 = 8 22 = 26 = 64 D-Wave), and observes that the number of qubits on
4 D-Wave’s processors has doubled every year for the
Year 4 24 = 16 22 = 28 = 256
past 10 years, in-line with Moore’s law

56
The number of qubits and the qubit quality both timing of when it will be possible to implement error
need to be significantly improved to develop a correction is uncertain, but the National Academy
fault-tolerant quantum computer. The National of Sciences, Engineering, and Medicine estimates it
Academy of Sciences, Engineering, and Medicine may occur by the early 2020s, which will be a critical
estimates that systems need to be about five orders milestone for the industry (National Academies of
of magnitude larger than current machines – current Sciences, Engineering, and Medicine, 2019).
NISQ-era machines have tens of physical qubits
(National Academies of Sciences, Engineering, Because of the error rates of current machines,
and Medicine, 2019). Furthermore, they estimate the number of logical qubits is at present
that error rates need to be reduced by two orders effectively zero; however, once error correction is
of magnitude (National Academies of Sciences, implemented, effectively creating logical qubits, a
Engineering, and Medicine, 2019). collection of such qubits will be able to carry out
an indefinitely long series of fault-tolerant qubit
Quantum error correction will play a key role in operations. Accordingly, to monitor progress in the
further reducing error rates; however, it cannot be near-term (before error correction is implemented),
implemented on current machines because many the industry can track the scale of systems (number
high-quality physical qubits are required, and many of physical qubits) and the average gate error rates.
gate operations must be performed. The worst gate In the longer term, the number of logical qubits will
in the system needs to have an error rate below become non-zero, and will be an appropriate metric
the threshold of 10-2 or 10-3 or better before error to track the achievable operational complexity of
correction can be applied (National Academies of the system.
Sciences, Engineering, and Medicine, 2019). The

Source: Maslov D, Nam Y, Kim J. An Outlook for Quantum Computing [Point of View]. Proceedings of the IEEE. 2019 Jan
1;107(1):5–10.

57
 METRICS FOR MONITORING PROGRESS engineers. While quantum science university
programs are established, other programs and
strategies to train a future quantum engineering
Near Term: Monitor Quantum error Long Term:
number of physical correction Monitor the
workforce are still emerging.
qubits and average implemented number of
gate error rates logical qubits Standards

Standardization efforts are also being framed and

As systems scale to more logical qubits (and developed to help coordinate and accelerate
necessarily more physical qubits and lower error progress. The development of standards could help
rates), new problem types are expected to become establish metrics for inter-device comparison and
addressable. For example, Gartner estimates could encourage supply chain development for an
that organic chemistry and portfolio optimization interoperable ecosystem, among other benefits.
applications will require systems on the order of a Multiple groups such as ANSI, ITU, and IEEE are
couple hundred logical qubits compared to the two engaging in early standardization efforts.
thousand or more required to crack RSA (Brisse,
Reynolds, & Horvath, 2019). Other applications, Software
such as route optimization and drug design, will
likely come in between. While this report has focused on the optics and
photonics components required to enable quantum
Research and development are rapidly advancing technologies, other hardware (e.g., cryogenic
but we are likely a decade or more away from coolers, control electronics) and software also
having a fault-tolerant quantum computer capable needs to be developed to enable the finished
of cracking RSA. In addition to the hardware systems. There is notable investment in software
improvements that are required, quantum startups such as Zapata Computing, 1QBit, and
algorithms and software stacks need to be Cambridge Quantum Computing, by the venture
developed to address applications and improve the capital community. Both the software stack and
efficiency of loading large data inputs into these the application-specific algorithms are under
machines. development. In fact, significant investments are
already being made in software and services, even
Quantum Workforce Development in advance of a functioning fault-tolerant quantum
computer.
In addition to the hardware and software supply chain
needs described above, education and workforce Societal Impacts
development are also required to enable progress.
Quantum technology is an interdisciplinary field While it is early to predict specific societal impacts
that engages physicists (theoretical, experimental, of quantum technology, it is safe to predict there
optical, etc.), computer scientists, mathematicians, might be potentially large impacts, both positive
chemists, engineers, material scientists, etc. While ones and those with risks or even negative impacts.
many university programs have been created Some scholars and observers are beginning now
to address the needs of the community, further to consider such issues (Vermaas, 2017). These
workforce development is needed to commercialize potential impacts should be actively considered as
and apply these technologies in real-world the technical ecosystem develops further.
applications.

For example, a critical transition needs to be made

to progress current lab-scale demonstrations to
scalable manufacturing, which will require quantum

58
CONCLUSION

The quantum technology market is still very early-stage. Volumes are modest and commercial timing is
uncertain. Nevertheless, there are opportunities for the optics and photonics community to participate in
this industry and supply the critical enabling components – sometimes compared to the “picks and shovels”
of the gold rush – to the well-funded R&D community in the near-term so that progress can continue.
Optics and photonics are core enabling technologies across many of the architectures and systems being
developed, and a measured investment today could yield future opportunities as the market for quantum
devices advances.

Of the three main categories of quantum technology, quantum computing receives the most media attention
but is also the most speculative and longer-term. Many large players such as IBM, Google, Intel, Microsoft,
Honeywell, etc. are investing in system development. However, the supply chain is still immature – many
of the necessary components are supplied by small, specialized players and startups (some of which are
sole suppliers). Furthermore, quantum computers are likely to remain highly specialized in the near term,
meaning that component volumes will remain low. Quantum sensing and communications are expected to
bring much earlier commercial success for OEMs and component suppliers alike.

HOW BIG IS THE MARKET?

2020 Market Size Expected Growth Potential

• Current sales primarily clocks, gravimeters, and SQUIDs

• Market forecasts vary widely – between US$0.5B (Yole)
SENSING
and >US$2B (INSIDE Quantum) by the end of the
US$414M decade
• Market penetration depends on ability to demonstrate
improvements versus incumbent technology

• Relatively less variation in today’s market size estimates

• Divergence between sources in market forecasts ranging
from US$785M (Yole) to over US$7B (INSIDE Quantum)
QKD
US$84M by end of decade
• Market uptake dependent on demonstrated tech.
advancements

• Current sales largely limited to quantum annealers

• Most agree that the future market size (once a large-
scale fault-tolerant computer is developed) will be
US$33M
COMPUTING comparable to the current supercomputer market which
is about US$50B
• Market forecasts vary likely due to uncertainty about the
pace of expected progress
Source: Yole Développement (2020)

59
Fortunately, many of the components required monitoring the following metrics and technological
for quantum sensors or communications networks milestones as leading indicators:
could also have utility in some types of quantum
computing architectures. While the winning • Commercialization of field-deployable quantum
approach to building a quantum computer has yet to repeaters will be a significant milestone for
emerge, optics and photonics component suppliers quantum communications, as it should have a
can de-risk investment in R&D by prioritizing substantial impact on potential transmission
platform technologies and components that: distance. Progress can be tracked by monitoring
the transmission distance and data rates
• can be used across multiple quantum application achieved in new experiments and commercial
categories: Some components are applicable installations.
to sensing, communications, and computing
markets. For example, lasers for atom-based • Implementation of error correction on a small
systems are required in specific types of quantum quantum computer will be a significant milestone
computing (e.g., neutral atom, trapped-ion) on the path to large-scale fault-tolerant
and quantum sensors based on cold atom quantum computing. Once error correction is
interferometry. Single-photon sources, single- implemented, the number of logical qubits will
photon detectors, and modulators are used in become non-zero. To monitor progress across
both discrete variable quantum communications the different groups, investors can track the
protocols and photonic quantum computing. number of logical qubits achieved by different
systems. Gartner estimates the first applications
• will be required regardless of which architecture (e.g., chemistry, portfolio, and/or route
wins: Regardless of which network architecture optimization) are likely to be addressable once
(e.g., superconducting versus ion trap) ~100-500 logical qubits are achieved. Longer
emerges as the preferred solution for quantum term, machines with numbers of logical qubits
computers, interconnects such as microwave- on the order of 2,000-3,000 may be required for
to-optical transducers and frequency converters breaking RSA encryption.
are likely to be required to connect the various
nodes and components in the network, all of • Due to the wide array of different sensing
which operate at different frequencies. applications, it is difficult to point to a single
milestone as a key landmark of progress. Instead,
• will have value in non-quantum markets: Some quantum sensors should be compared to classical
components required for quantum technology incumbents based on their performance, cost,
will also have utility in other markets that and SWAP. Commercialization of quantum
may demand higher volumes. For example, sensors for GPS-free navigation would be a
single-photon detectors can be used in low- notable technological milestone.
light imaging for life sciences and metrology
applications. Integrated photonic chips and While the exact timing of these advancements is
couplers also have broad utility in larger markets, not definite, most experts currently project that
including classical telecom. significant progress will be made in the next 5-10
years. Monitoring the progress toward these goals
As system performance improves and over the next decade will provide a much clearer
implementation uncertainties are resolved, the picture. Despite the highly uncertain nature of the
community may begin to coalesce around specific market for quantum technology today, the scale of
devices for specific applications. As this change ongoing development activity and the potential for
occurs, the optics and photonics community can future disruption justifies measured investments to
scale and refine investment further. We recommend understand what opportunities could materialize in
the near- and long-term.

60
/////////////// MILESTONES AND METRICS
Current State Future Vision

Integration.
Gravimeters & Large- Miniaturized
SENSING Format Atomic Clocks Stability, Sensors
SWAP-C

Quantum Repeater,
Point-to-Point Quantum
COMMS Distance,
QKD Links Network
Key Rate

Annealers & NISQ Era Gate-Based Error Correction & Fault-Tolerant

COMPUTING
Computers Logical Qubits Quantum Computer

Advancing Progress: What’s Required

FUNDING HARDWARE
Sustained public and private Performance improvements, cost reduction, supply
investment to support R&D / scaling chain development for components

WORKFORCE SOFTWARE
Education to fill the talent gap and Error correction, algorithm development,
build the future quantum workforce software stack, etc.

61
QUESTIONS? CONTACT US!

For more information about the report, please contact the project team:

Tom Hausken, PhD Kasey O’Malley

Senior Advisor Senior Consultant
The Optical Society (OSA) and OSA Industry Newry Corp.
Development Associates (OIDA) [email protected]
[email protected]
Ellen Young
Elise Fritz-Gage Senior Vice President
Senior Manager, Industry Programs & Events Newry Corp.
The Optical Society (OSA) and OSA Industry [email protected]
Development Associates (OIDA)
[email protected] Andrew Gartley
Consultant
Newry Corp.
[email protected]

Coming in June 2020!

Quantum 2.0 is a new OSA
conference focusing on
the development and use
of many-body quantum
superposition, entanglement,
and measurement to
advance science and
technology.

62
BIBLIOGRAPHY

Accenture. (2020). Biogen, 1QBIT, and Accenture: Pioneering Quantum Computing in R&D. Retrieved
from https://www.accenture.com/us-en/success-biogen-quantum-computing-advance-drug-
discovery

Accenture Labs. (2017). Innovating with Quantum Computing. Retrieved from https://www.accenture.
com/t00010101t000000__w__/br-pt/_acnmedia/pdf-45/accenture-innovating-quantum-
computing-novo.pdf

Allgaier, M., Ansari, V., Sansoni, L. et al. (2017). Highly efficient frequency conversion with bandwidth
compression of quantum light. Nature Communications, 8, 14288.

Awschalom, D., Berggren, K. K., Bernien, et. al. (2019). Development of Quantum InterConnects for Next-
Generation Information Technologies. arXiv preprint arXiv:1912.06642.

Ball, P. (2018, January 24). The Era of Quantum Computing Is Here. Outlook: Cloudy. Quanta Magazine.
Retrieved from https://www.quantamagazine.org/the-era-of-quantum-computing-is-here-
outlook-cloudy-20180124/

Battersby, S. (2019, August 20). Core Concept: Quantum sensors probe uncharted territories, from
Earth’s crust to the human brain. PNAS 116 (34), pp. 16663-16665.

Boaron, A., Boso, G., Rusca, D., Vulliez, C., Autebert, C., Caloz, M., . . . Zbinden, H. (2018). Secure
Quantum Key Distribution over 421 km of Optical Fiber. Phys. Rev. Lett., 121, 190502.

Bogdanov, S., Shalaginov, M. Y., Boltasseva, A., & Shalaev, V. M. (2017). Material platforms for integrated
quantum photonics. Optical Materials Express, 7(1), 111-132.

Bourzac, K. (2019, April 12). To upgrade quantum computers, researchers look to materials science.
Chemical & Engineering News, 97(15). Retrieved from https://cen.acs.org/materials/electronic-
materials/upgrade-quantum-computers-researchers-look/97/i15

Brisse, M., Reynolds, M., & Horvath, M. (2019). Webinar: Taking the Quantum Leap - Fact, Fiction, or
Fantasy. Gartner.

Carleo, G., Cirac, I., Cranmer, K., Daudet, L., Schuld, M., Tishby, N., . . . Zdeborova, L. (2019). Machine
learning and the physical sciences. Rev. Mod. Phys., 91, 045002.

Cartlidge, E. (2016, October). Quantum Computing: How Close Are We? Optics & Photonics News, 30-
37.

Cho, Y.-W., Campbell, G., Everett, J., Bernu, J., Higginbottom, D., Cao, M., . . . Buchler, B. (2016). Highly
efficient optical quantum memory with long coherence time in cold atoms. Optica, 3(1), 100-107.

Chunnilall, C. J., Degiovanni, I. P., Kuck, S., Muller, I., & Sinclair, A. G. (2014). Metrology of single-photon
sources and detectors: a review. Optical Engineering, 53(8), 081910.

Courtland, R. (2016, October 26). China’s 2,000-km Quantum Link Is Almost Complete. Retrieved from
IEEE Spectrum: https://spectrum.ieee.org/telecom/security/chinas-2000km-quantum-link-is-
almost-complete

63
Diamanti, E., Lo, H.-K., Bing, Q., & Yuan, Z. (2016). Practical challenges in quantum key distribution. npj
Quantum Information, 2, 16025.

D-Wave Systems Inc. (2019). Quantum Applications. Retrieved from https://www.dwavesys.com/sites/

default/files/3_D-Wave-Qubits-Seminar-Applications_ajsm_V0.1.pdf

D-Wave Systems, Inc. (n.d.). Introduction to the D-Wave Quantum Hardware. Retrieved from D-Wave
Systems Tutorials: https://www.dwavesys.com/tutorials/background-reading-series/introduction-
d-wave-quantum-hardware

Eisaman, M. D., Fan, J., Migdall, A., & Polyakov, S. V. (2011). Invited Review Article: Single-photon sources
and detectors. Review of Scientific Instruments, 82, 071101.

European Cooperation in Science & Technology, Tematys. (2019). Market Research Study in Nanoscale
Quantum Optics. Retrieved from https://www.cost-nqo.eu/wp-content/uploads/2019/04/COST-
NQO-Market-Research-Study-web.pdf

Finke, D. (2019). A Quantum Computing Use Case Roadmap from IBM. Retrieved from Quantum
Computing Report: https://quantumcomputingreport.com/our-take/a-quantum-computing-
application-roadmap-from-ibm/

Finke, D. (January, 9 2019). Qubit Count. Retrieved from Quantum Computing Report: https://
quantumcomputingreport.com/scorecards/qubit-count/

Flamini, F., Spagnolo, N., & Sciarrino, F. (2018). Photonic quantum information processing: a review.
Reports on Progress in Physics, 82(1), 016001.

Gehring, H. G., Blaicher, M., Hartmann, W., Varytis, P., Busch, K., Wegener, M., & Pernice, W. P. (2019). Low-
loss fiber-to-chip couplers with ultrawide optical bandwidth. APL Photonics, 4, 010801.

Gerbert, P., & Rueß, F. (2018, November 15). The Next Decade in Quantum Computing—and How to Play.
Retrieved from BCG: https://www.bcg.com/en-us/publications/2018/next-decade-quantum-
computing-how-play.aspx

Gibney, E. (2019). The quantum gold rush. Nature, 574, 22-24.

Giles, M. (2018, December 22). MIT Technology Review. Retrieved from https://www.technologyreview.
com/f/612679/president-trump-has-signed-a-12-billon-law-to-boost-us-quantum-tech/

Gooch & Housego, Milner Strategic Marketing, & University of Bristol. (2018). The UK Market for
Quantum Enabling Photon Sources: 2018-2022. Retrieved from https://gandh.com/wp-content/
uploads/2018/05/GH-Market-Report-UK-Market-for-Quantum-Enabling-Photon-Sources-2018-
2022-Report.pdf

Guo, J., Feng, X., Yang, P., Yu, Z., Chen, L. Q., Yuan, C., & Zhang, W. (2019). High-performance Raman
quantum memory with optimal control in room temperature atoms. Nature Communications, 10,
148.

64
Hartnett, K. (2019, June 18). A New Law to Describe Quantum Computing’s Rise? Retrieved from
Quanta Magazine: https://www.quantamagazine.org/does-nevens-law-describe-quantum-
computings-rise-20190618/

Herman, A. (2019, April 3). The Executive’s Guide to Quantum Computing and Quantum-secure
Cybersecurity. Retrieved from Hudson Institute: https://www.hudson.org/research/14930-the-
executive-s-guide-to-quantum-computing-and-quantum-secure-cybersecurity

Heshami, K., England, D. G., Humphreys, P. C., Bustard, P. J., Acosta, V. M., Nunn, J., & Sussman, B. J.
(2016). Quantum memories: emerging applications and recent advances. Journal of Modern
Optics, 63(20), 2005-2028.

Holzman, I., & Ivry, Y. (2019). Superconducting Nanowires for Single-Photon Detection: Progress,
Challenges, and Opportunities. Advanced Quantum Technologies, 2(3-4), 1800058.

Hussain, M., Petrasiunas, M., Bentley, C., Taylor, R., Carvalho, A., Hope , J., . . . Kielpinski, D. (2016).
Ultrafast, high repetition rate, ultraviolet, fiber-laser-based source: application towards
Yb⁺ fast quantum-logic. Opt Express, 24(15), 16638-48.

IBM. (2019, September 10). IBM and Fraunhofer Join Forces on Quantum Computing Initiative
for Germany. Retrieved from IBM News Room: https://newsroom.ibm.com/2019-09-10-IBM-
and-Fraunhofer-Join-Forces-on-Quantum-Computing-Initiative-for-Germany#:~:text=The%20
German%20Government%20will%20invest,European%20hub%20for%20quantum%20
computing.

ID Quantique. (2017, November 23). IDQ Celebrates 10-Year Anniversary of the World’s First Real-Life
Quantum Cryptography Installation. Retrieved from https://www.idquantique.com/idq-
celebrates-10-year-anniversary-of-the-worlds-first-real-life-quantum-cryptography-installation/

InfiniQuant. (n.d.). Satellite Quantum Communication. Retrieved from InfiniQuant: http://infiniquant.com/

front-page/satellite-quantum-communication/

INSIDE Quantum Technology. (2019, January 14). Quantum Sensors: Ten-Year Market Projections.
Retrieved from https://www.insidequantumtechnology.com/wp-content/uploads/2019/03/
qsensors-for-chrissie-1.pdf

INSIDE Quantum Technology. (2019, June). The Future of the Quantum Internet: A Commercialization
Perspective. Presentation at ITU Workshop on Quantum Information Technology for Networks.
Retrieved from https://www.insidequantumtechnology.com/wp-content/uploads/2019/06/itu-t-
slides-final-the-future-of-the-quantum-internet.pdf

Kahn, J. (2019, January 14). The Super-Secure Quantum Cable Hiding in the Holland Tunnel. Retrieved
from Bloomberg Businessweek: https://www.bloomberg.com/news/articles/2019-01-14/the-
super-secure-quantum-cable-hiding-in-the-holland-tunnel

Kania, E. B., & Costello, J. K. (2018). Quantum Hegemony?: China’s Ambitions and The Challenge to U.S.
Innovation Leadership. Center for New American Security.

65
Khan, I., Heim, B., Neuzner, A., & Marquardt, C. (2018, February). Satellite-Based QKD. Optics & Photonics
News, 26-33.

Lambert, N. J., Ruedo, A., Sedlmeir, F., & Schwefel, H. G. (2020). Coherent conversion between
microwave and optical photons -- an overview of physical implementations. Advanced Quantum
Technologies, 3(1900077).

Lance, A., & Leiseboer, J. (2014). What is Quantum Key Distribution (QKD)? White Paper #1. Quintessence
Labs. Retrieved from https://www.zeiss.com/content/dam/Corporate/innovation_and_
technology/downloads/zeiss-symposium_whitepaper_qms_qei.pdf

Li, Z.-D., Zhang, R., Yin, X.-F., Liu, L.-Z., Hu, Y., Fang, Y.-Q., . . . Pan, J.-W. (2019). Experimental quantum
repeater without quantum memory. Nature Photonics, 13, 644-648.

Lo, H.-K., Curty, M., & Qi, B. (2012). Measurement-Device-Independent Quantum Key Distribution.
Physical Review Letters, 108(13), 130503.

Ma, L., Tang, X., & Slattery, O. (2019). Optical quantum memory applications in quantum communication.
SPIE Optical Engineering + Applications, 1113401. San Diego.

Mailloux, L. O., Grimaila, M. R., Hodson, D. D., Baumgartner, G., & McLaughlin, C. (2015, Jan.-Feb.).
Performance Evaluations of Quantum Key Distribution System Architectures. IEEE Security &
Privacy, 13, pp. 30-40.

Malcolm, G., & Hempler, N. (2019, May 27). Shedding Light on Quantum Computing. Retrieved from
Novus Light: https://www.novuslight.com/shedding-light-on-quantum-computing_N9226.html

Manojlovic, V. (2017, May 26). Introduction to the quantum internet. Retrieved from RIPE NCC: https://
labs.ripe.net/Members/becha/introduction-to-the-quantum-internet

Marchetti, R., Lacava, C., Carroll, L., Gradkowski, K., & Minzioni, P. (2019). Coupling strategies for silicon
photonics integrated chips. Photonics Research, 7(2), 201-239.

Maslov, D., Nam, Y., & Kim, J. (2019, January). An Outlook for Quantum Computing. Proceedings of the
IEEE, 107(1), pp. 5-10.

Mosca, M. (2018, September / October). Cybersecurity in an Era with Quantum Computers: Will We Be
Ready? IEEE Security & Privacy, 16(5), 38-41.

Munro, E. (2018, July 22). Quantum sensing. Medium. Retrieved from https://medium.com/@quantum_
wa/quantum-sensing-f33643d098bb

National Academies of Sciences, Engineering, and Medicine. (2019). Quantum Computing: Progress and
Prospects. Washington, DC: The National Academies Press.

Palmer, J. (2017, March 9). Quantum technology is beginning to come into its own. The Economist.

66
Peach, M. (2019, October 18). QuTech and Fraunhofer ILT launch ‘ICON’ quantum optical network
project to put quantum into the fibre optic network. Retrieved from Optical Connections: https://
opticalconnectionsnews.com/2019/10/qutech-and-fraunhofer-ilt-launch-icon-quantum-optical-
network-project-to-put-quantum-into-the-fibre-optic-network/

Popkin, G. (2016, December 1). Scientists are close to building a quantum computer that can beat a
conventional one. Science Magazine. Retrieved from https://www.sciencemag.org/
news/2016/12/scientists-are-close-building-quantum-computer-can-beat-conventional-one

Preskill, J. (2018). Quantum Computing in the NISW era and beyond. Quantum, 2, 79.

Quantum Business. (2018, January 26). Japan Intends To Be At the Forefront of the Quantum Leap.
Retrieved from Quantum Business: https://quantumbusiness.org/japan-intends-forefront-
quantum-leap/

Quantum Delta Nederland. (2019, September 16). National Agenda on Quantum Technology. Retrieved
from https://www.nwo.nl/en/news-and-events/news/2019/09/national-agenda-on-quantum-
technology-the-netherlands-as-an-international-centre-for-quantum-technology.html

Quantum Flagship. (2019, July 31). Strategic Research Agenda.

Roberson, T. M., & White, A. G. (2019). Charting the Australian quantum landscape. Quantum Science
and Technology, 4, 020505.

Sinha, U., Sahoo, S. N., Singh, A., Joarder, K., Chatterjee, R., & Chakraborti, S. (2019). Single-Photon
Sources. Optics & Photonics News, 32-39.

SK Telecom. (2019, March 18). SK Telecom Continues to Arm its 5G Network with Quantum Cryptography
Technologies. Retrieved from https://www.sktelecom.com/en/press/press_detail.do?page.
page=1&idx=1385&page.type=all&page.keyword=

Slussarenko, S., & Pryde, G. J. (2019). Photonic quantum information processing: A concise review.
Applied Physics Reviews, 6, 041303.

Soltani, M., Soref, R., Palacios, T., & Englund, D. (2016). AlGaN/AlN integrated photonics platform for the
ultraviolet and visible spectral range. Optics Express, 24(22), 25415-25423.

Stewart, D. (2018). Quantum computers: the next super computers, but not the next laptops. Deloitte
Technology, Media, and Telecommunications Predictions 2019. Retrieved from https://www2.
deloitte.com/content/dam/Deloitte/pe/Documents/technology-media-telecommunications/
Quantum%20computers.pdf

Sussman, B., Corkum, P., Blais, A., Cory, D., & Damascelli, A. (2019). Quantum Canada. Quantum Science
and Technology, 4, 020503.

Temme, K., & Gambetta, J. (2019, March 13). Researchers Put Machine Learning on Path to Quantum
Advantage. Retrieved from IBM Research Blog: https://www.ibm.com/blogs/research/2019/03/
machine-learning-quantum-advantage/

67
TU Delft. (2019, June 2). Quantum Internet: The internet’s next big step.

UK Government Office for Science. (2016). The quantum age: Technological opportunities. Retrieved
from https://www.gov.uk/government/publications/quantum-technologies-blackett-review

UK National Quantum Technology Hub: Sensors and Metrology. (2018). Outline of quantum sensors
technology roadmap. Retrieved from https://www.quantumsensors.org/wp-content/
uploads/2019/05/Outline-publication_A4_Gotham_v1.pdf

UK Research and Innovation. (2019, June 13). £1 billion investment makes UK a frontrunner in quantum
technologies. Retrieved from GOV.UK: https://www.gov.uk/government/news/1-billion-
investment-makes-uk-a-frontrunner-in-quantum-technologies

Vermaas, P. E. (2017, December). Special Issue: The Societal Impact of the Emerging Quantum
Technologies. Ethics and Information Technology, 19(4), 241-306.

Volkswagen. (2018, June 7). Volkswagen tests quantum computing in battery research. Retrieved from
https://www.volkswagenag.com/en/news/2018/06/volkswagen-tests-quantum-computing-in-
battery-research.html

Wehner, S., Elkouss, D., & Hanson, R. (2018). Quantum internet: A vision for the road ahead. Science,
362(6412).

Wicht, A., Krutzik, M., & Thoss, A. (2018, January 18). Quantum Technology: Quantum sensing is
gaining (s)pace. LaserFocusWorld. Retrieved from https://www.laserfocusworld.com/lasers-
sources/article/16555248/quantum-technology-quantum-sensing-is-gaining-space

Widmer, L., & Pauchard, A. (2006, May). Quantum detection: photon counters in comparison. Photonik,
pp. 40-43.

Yang, S. W. (2016). High-fidelity transfer and storage of photon states in a single nuclear spin. Nature
Photonics, 10, 507-511.

Yin, H.-L., Chen, T.-Y., Yu, Z.-W., Liu, H., You, L.-X., Zhou, Y.-H., . . . Pan, J.-W. (2016). Measurement-Device-
Independent Quantum Key Distribution Over a 404 km Optical Fiber. Physical Review Letters,
117(19), 190501.

Yirka, B. (2019, November 8). A new way to measure gravity: Using floating atoms. Retrieved from Phys.
org: https://m.phys.org/news/2019-11-gravity-atoms.html

Yole Développement. (2020, January). Quantum Technologies Flyer: Market and Technology Report
2020. Retrieved from https://yole-i-micronews-com.osu.eu-west-2.outscale.com/
uploads/2020/01/YDR20062-Quantum-Technologies-2020-Flyer-Yole-D%C3%A9veloppement.
pdf

Zeiss. (2018). White Paper on Quantum Computing and Quantum Communication. Retrieved from
https://www.zeiss.com/content/dam/Corporate/innovation_and_technology/downloads/zeiss-
symposium_whitepaper_qc_qcom.pdf

68
Zeiss. (2018). White Paper on Quantum Metrology & Sensing and Quantum Enhanced Imaging. Retrieved
from https://www.zeiss.com/content/dam/Corporate/innovation_and_technology/downloads/
zeiss-symposium_whitepaper_qms_qei.pdf

Zhao, T., Chen, Y., Yu, Y., Li, Q., Davanco, M., & Liu, J. (2019). Advanced Technologies for Quantum
Photonic Devices Based on Epitaxial Quantum Dots. Advanced Quantum Technologies. Retrieved
from https://onlinelibrary.wiley.com/doi/full/10.1002/qute.201900034

Zheltikov, A. M. (2018). Enhanced-contrast optical readout in ultrafast broadband Raman quantum

memories. Sci Rep, 8, 13774.

69