WEF量子计算的现状:构建量子经济 2022
WEF量子计算的现状:构建量子经济 2022
Contents
Preface 3
Executive summary 4
Introduction 5
3 State of technology 26
Conclusion 42
Contributors 43
Appendix 44
Endnotes 45
Disclaimer
This document is published by the
World Economic Forum as a contribution
to a project, insight area or interaction.
The findings, interpretations and
conclusions expressed herein are a result
of a collaborative process facilitated and
endorsed by the World Economic Forum
but whose results do not necessarily
represent the views of the World Economic
Forum, nor the entirety of its Members,
Partners or other stakeholders.
Preface
Derek O’Halloran
Head, Shaping the Future of Digital
Economy and New Value Creation;
Member, Executive Committee,
World Economic Forum
This report focuses on the computing aspect, At the same time, today’s imperfect quantum
where the unique capabilities of quantum computers are suitable to run a subset of
computers open the potential for tackling applications where a precise answer is not required
complex computation problems that, as a result and seeing trends or a likely direction is more
of fundamental limitations in scaling classical important. This, coupled with cloud access now
computation, would be intractable with classical offered by many quantum computing providers,
computing methods and unlocking new possibilities has allowed many more organizations to start
that have not yet been considered. Consequently, experimenting with quantum applications. Given
many expect that the impact of quantum computing rapid technological developments, it is only a matter
will be as foundationally transformative as the onset of time before an application that provides a real
of classical computing in the mid-20th century. It advantage with the available quantum technology
also comes with risks, as quantum computers are is found. When this occurs, and what application
expected to break the encryption used to secure can make best use of the available computational
modern digital communications, blockchain and resources is the name of the game.
some cryptocurrencies.
While these activities receive a lot of attention,
It is important to note that quantum computing the current quantum computing achievements
is expected to work best across three specific and future development rely on several enablers:
domains of research and industry, with significant workforce readiness, standardization and policy.
economic, environmental and societal opportunities Workforce availability is the critical bottleneck.
associated with them:
Despite uncertainty over when quantum
1. Molecular simulation and discovery in materials computers will be ready at scale, governments
science and biology and businesses must act now, as quantum
security risks and business opportunities cannot
2. Optimization and risk management in be ignored. This is a unique moment in modern
complex systems history where everyone can prepare for the
technology as it is being shaped and matured.
3. A bi-directional impact on existing technology Governments and academia can further work
areas such as AI, security and blockchain. on scaling quantum workforce programmes
and building national ecosystems, incentivizing
All in all, quantum computing is positioned to partnerships. For example, an excellent first
complement classical computing and be applied step for businesses is to understand quantum
to certain tasks outside the reach of today’s computing’s impact on business and industry,
supercomputers, carving out its powerful assess their quantum readiness and formulate a
computational niche that didn’t exist before, quantum strategy, build internal capabilities, and
eventually expanding modern research and align with top management and policy-makers
development and business horizons. on critical focus areas.
1 0
z
Superposition (50%) 1
0
Overlay of different states
or
x y
0 1
Can be only in one state Can be in a superposition state, by being in one of
at the same time, 0 or 1 multiple states of 0 and 1 at the same time
Below are some key terms and definitions that Quantum error correction (QEC): A
will become familiar over the course of 6 computing technique for dealing with errors
this report. in quantum computers that exploits encoding
across extra (large) number of qubits to
reduce errors rates.
Quantum technologies: Any technology
1 that uses the principles of quantum
physics, including quantum computing,
medical devices, highly-sensitive sensors,
7 Noisy intermediate-scale quantum (NISQ)
devices: NISQ is a commonly used term to
secure communications, atomic clocks, describe today’s quantum computers, which
etc. Quantum computing is a subset of are “noisy” (i.e. prone to errors, which can
quantum technologies. accumulate and lead to incorrect computation)
and are of intermediate-scale (i.e. these
devices have tens to hundreds of qubits).
Quantum computing (QC): A new
2 approach to computation that exploits the
nature of storing and manipulating quantum Quantum circuit: A sequence of quantum
information. 8 operations applied on multiple qubits. It is
an ordered sequence of quantum gates,
measurements and resets, forming the basis
Qubit: A quantum bit, a unit of information of quantum algorithms.
3 storage in quantum computing
fundamentally different to a classical “bit”
that can hold a linear superposition of Quantum algorithm: A collection of quantum
states, meaning it can be zero and one at 9 circuits to run on a quantum computer to find
the same time, but upon measurement, a solution to a problem. Quantum algorithms
reveal either zero or one with a defined are the engines of quantum applications. One
probability distribution. algorithm can be the basis of multiple
applications in different industries.
Quantum computing
activities can be seen
in every continent, with
major flagship initiatives
materializing across
most G20 countries.
$35.5 billion
Public and private investments totalled
$35.5 billion by 2022 across a range of
quantum technologies.
Countries with The development of quantum computing The hefty mix of public and private investments has
leading quantum technologies today is a global endeavour. sparked a diverse constellation of companies to
R&D clusters All continents are home to companies and develop quantum computers or key building blocks. At
from around governments actively supporting the creation of the beginning of 2022, 46 companies worldwide were
the globe have new, quantum-powered computing solutions. actively developing quantum computing hardware.4
Countries with leading quantum R&D clusters from A growing number of start-ups are entering the scene
made strategic
around the globe have made strategic investments and attracting funding from investors worldwide. As
investments to
to capture part of the future quantum computing a result, today’s early-stage quantum computers
capture part of the supply chain and create strategic and independent work with a wide variety of hardware platforms in
future quantum access to future capabilities. different stages of development. These include
computing supply superconducting qubits, neutral atoms, trapped
chain. These multi-billion-dollar investments fuel ions, photonic qubits and silicon-based qubits.
technological advances. According to the
latest research, public investments in quantum These different quantum computing flavours reflect
technologies exceed $30 billion (see Figure 1). a dynamic and rapidly evolving technology sector.
Private investments for quantum technologies Significant milestones have been reported from
added $3.2 billion in 2021 alone and over $5.5 different corners of the planet: the first quantum
billion in the past decade.3 It is important to note processor was made available to everyone
that these numbers are underestimations since not through the cloud access in 2016,5 in 2019 the
all public investment is captured due to national achievement of quantum supremacy was claimed,6
security concerns and what is tracked and reported whereby the quantum processor realized in a
in various regions. While some large companies matter of minutes a complex calculation that would
have private and public funds, a few heavily take the world’s most powerful computer system at
support their quantum efforts from corporate R&D the time two days to process. Improvements in the
funds. As with public investments, corporate R&D error rates for qubit control and reliable read-out
funds are typically not captured by the private have continued to be a part of the R&D landscape,
investments described above. as well as critical breakthroughs in scaling.
Denmark Sweden
DKK 230 million SEK 1.6 billion
= $34 million = $160 million Russia
RUB 50 billion = $663 million
Netherlands Finland
€765 million €24 million
= $904 million = $27 million China
$15 billion
United Kingdom
£1 billion
Canada = $1.3 billion South Korea
Israel KRW 44.5 billion
CAD 1.37 billion = $1.1 billion ILS 1.2 billion
France = $40 million
= $380 million
€1.8 billion
Japan
US National Quantum Initiative = $2.2b
JPY 80 billion = $700 million
$1.2 billion India
Spain INR 73 billion = $1 billion Taiwan
€60 million TWD 8 billion = $282 million
= $67 million Thailand
THB 200 million
= $6 million Singapore
Germany SGD 150 million = $109 million
Global effort 2022: €2.6 billion = $3.1 billion
Hungary
HUF 3.5 billion = $11 million New Zealand
NZD 36.75 million = $20.9 million
Asia, North America, Europe and Australia have efforts on high-performance computing. There
very different innovation ecosystems. Consequently, are also initiatives in addition to existing research
national governments follow different pathways centres and clusters of excellence in quantum
in the quantum computing journey. The United technologies, e.g. Quantum Delta NL in the
States established a national programme in 2019, Netherlands, a dedicated legal entity that governs
the National Quantum Initiative Act,7 to support all related public investments in technology.
the development of all quantum technologies. This
includes the founding of the Quantum Economic In line with these investments, the European
Development Consortium (QED-C) to support the industry has also started to coalesce to accelerate
development of a quantum supply chain with the goal the development of commercial quantum solutions.
and mission to support the future quantum industry. One example is the European Quantum Industry
Canada and its provinces continue to work to evolve Consortium (QuIC). This pan-European organization
their national- and provincial-level strategies. brings together start-ups, small and medium
enterprises (SMEs), large companies, investors,
In Europe, national and regional initiatives have research and technology organisations, and other
been launched to spearhead the development associations. At national levels, several other
of quantum computing solutions. The UK, the industry associations have also been formed, such
Netherlands, Germany and France have developed as Le Lab Quantique (France), the Danish Quantum
national quantum programmes and strategies, Community (Denmark), the Finnish Institute Q and
with total public support of over $7 billion.8 The UK Quantum (United Kingdom).
European Commission has also established a
separate $1.1 billion research and innovation Asia has had a long and growing effort across
National initiative known as the EU Quantum Flagship, which the quantum landscape. Singapore’s focused
governments follow is dedicated to developing and commercializing effort in quantum information dates back to the
different pathways quantum technologies in the European Union. early 2000s. This was followed by a significant
in the quantum Within the EU, quantum computing is prominent effort in China that initially focused on quantum
computing journey. and developed in collaboration with European communications but now includes a large focus on
FIGURE 3 China and European Union lead significantly on public funding for quantum computing
China 15.3
Japan 1.8
India 1.0
Canada 1.0
Russia 0.7
Israel 0.5
Singapore 0.3
Australia 0.2
Other 0.1
1.7
Germany France European Union Netherlands Sweden Others 2.6
Source: McKinsey & Company, adapted from Johnny Kung and Muriam Fancy, 2021
Private financing has enabled the rise in quantum 196 start-ups globally have been established.12
computing start-ups. Figure 4 shows the rise in quantum computing
start-ups globally since 2015. The growth in private
The first commercial special-purpose quantum companies has been enabled by private financing.
computing company was founded in 1999, and Figure 5 shows that the amount of private financing
thirteen years later, the first quantum computing and the number of quantum computing start-ups
software company was founded.11 Since then, have grown rapidly in tandem.
23
5 1 1
19 53
59 4 9
15 7
1
6
1 18
3
7
2 1 2
2021 2015
Source: McKinsey & Company, adapted from PitchBook, 2021, World Economic Forum analysis
~800
800 40
2/3
679
Deals completed
$800
529
400 20
249 million
226
Equity investments could
have reached a single-year
200 132 10 record in 2021 (estimation)
200 205
92 143
72 89
88
14 70
58
73
70
%
10
0 0
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
In 2020, more than 90% of this investment went to capital investments needed to develop hardware
hardware players. Yet, the number of software and and the expectation of many that, as in the current
algorithm start-ups is growing faster than hardware ICT economy, there can be a lot of value created in
start-ups. This can be explained by the massive the application domain.
Quantum
computing
industries Total
Number of players
100
33
>
suppliers
72
Largely not specific
to quantum
computer hardware; 228
there are 38
QC-focused
components 39
suppliers that figure
into the overall
company count
46
4% 73% 14% 7% 2%
Share of
start-up funding
Source: McKinsey & Company, adapted from CaptalIQ, Crunchbase, PitchBook, 2022
FIGURE 7 As capital investment in quantum computing rises, more money is focused on software
$440 million
invested in software
1,500
1,357
100
1,000
782
Deals completed
Invested equity ($ millions)
798 50
$764 million
invested in software from
500
46 6
73 20 325
175
239
609
609
0 0
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Deal size
Company and HQ Deal date Select investors
($ millions)
SPAC dMY
New York Stock
IonQ Technology Group, 8 March 2021 $636 million
Exchange
Inc III
SPAC Supernova
Rigetti 13 NASDAQ Partners Acquisition 6 October 2021 $261.75 million
Company II
Centricus Acquisition
Arqit 14 NASDAQ 12 May 2021 $70 million
Corp
Source: SEC Filings, World IonQ was the first quantum computing company earlier deal, on 25 May 20021 Quantum Benchmark
Economic Forum, Global to do this, becoming the “first publicly traded, was acquired by Keysight Technologies.18
Future Council on Quantum pure-play quantum computing company”.16 They
Computing analysis
chose were followed shortly by Rigetti, Arqit and Recent mergers and acquisitions have also been
D-Wave, although it remains to be seen whether observed in Europe. In 2021 Zurich Instruments
this strategy stays as all four companies are trading was acquired by the German technology group
below the initial pricing points, with some losing up Rohde & Schwarz.19 In January 2022, the Dutch
to 50% value.17 Qu&Co merged with Pasqal.20 A few months
later, the Danish QDevil became part of Quantum
Following the Honeywell investment in Cambridge Machines.21 Additionally, many young start-ups
Quantum Computing, in June 2021, Honeywell came out of stealth and experienced steep growth
Quantum Systems merged with Cambridge Quantum trajectories fuelled by venture capital and European
Systems to form Quantinuum in late 2021. In an Investment Bank investments.
According to popular perception, quantum with digital computing. Given the nature of how
computers will be faster computers. This is not computation is consumed, it will naturally co-exist
necessarily true, however, as quantum computing with rather than replace classical computing.
will have strengths and weaknesses compared
Classical computers are expected to stay data), or do not provide any meaningful advantage
alongside quantum ones, to support quantum over existing computers (e.g. browsing the
computing processing workflow (e.g. running internet, reading emails, etc.). In other words,
quantum circuits) in hybrid data centres, and to if an application does not require quantum
continue powering daily tasks for which quantum circuits to run, it will likely continue to be run by
computers are either not suited (e.g. copying classical computers.
The rebuild of computing from “bits up” opens corresponding application of this ability is breaking
the door to tackling new types of problems that RSA encryption, which is the foundation of the
are not feasible through classical computing. majority of today’s secure data-transfer methods.
Quantum computing can tackle certain classes
of mathematical problems, which correspond to Key types of mathematical and physical problems
specific problem statements in the world. that quantum computers are expected to address
exceptionally well include quantum simulation,
For example, quantum computers have the optimization, quantum linear algebra and prime
provable potential to factor large numbers, a factorization. The ability to solve these abstract
mathematical problem so challenging that it is used problems (one by one or in combination) opens
to encrypt data transfers, exponentially faster than real-world opportunities in three broad areas
classical computers. The most straightforward (see Table 3):
Existing technology
Application domains Materials science and biology Complex systems
and research
Affected industries Energy, food and agriculture, Finance, transport and logistics, Industries with intense use of AI,
manufacturing, chemicals, industries with complex products blockchain or high-performance
medicine. (aviation, automotive, etc.). computing (HPC) in general,
energy and materials industries,
digital communications, defence
and security.
Quantum Discovery and design of new Management and optimization Impact on existing technologies
computing use molecules and materials, of sophisticated systems with such as AI, blockchain, as well
cases affecting multiple fields: advanced a large number of variables or as our scientific capabilities.
materials development, drug unknowns, from highly complex
design, crops and fertilizers, scheduling, logistical and supply
green hydrogen catalysts, chain challenges to modelling
batteries, chemistry. financial portfolios and risk
profiles to assessing national
defence strategies.
Social and Reduced energy consumption, Reduced energy consumption Breaks current cryptography,
environmental carbon capture, efficient and emissions across potential stronger
impacts materials and processes, more global networks, circular cryptography with increased
robust and nature-friendly crop business models. privacy and security.
variety, accelerated disease
and response discovery, Accelerates exploration and
personalized medicine. discovery in fundamental
science research.
Illustrative Molecules with the right Optimizing empty shipping The ability to accelerate training
examples attributes to sequester carbon container space yielding of machine learning algorithms.
at scale. both environmental and
economic gains. Breaking RSA and
More naturally resilient grains to cryptocurrencies encryption.
improve food production while Improving real-time customer
avoiding monocultures. credit scoring. Contributing to our fundamental
understanding of the quantum
behaviour of nature.
Underlying
Quantum simulation, optimization, quantum linear algebra and prime factorization.
quantum problems
Source: World Economic Quantum computing is based on the ability to 1) advantage has yet to be realized. The current
Forum, Global Future create and 2) control quantum circuits. To be status is reviewed in further depth in the “State of
Council on Quantum effective in real-world deployment, this needs Technology” section. As such, the applications in
Computing
to happen 3) at scale and with high quality section 2.2 are based on a shared understanding
and speed. While different quantum computing of where this technology is naturally suited to
development paths are at various stages of have a major impact – based on the nature of
maturity, reaching appropriate scale, quality the problem and the natural capabilities of
and speed to demonstrate impactful quantum quantum computing.
– Design novel batteries and energy storage – Novel crop-protection chemicals with the
systems, accelerating quantum simulation of desired characteristics are more likely to
electrolyte, anode and cathode materials, and be designed using quantum simulation and
choosing the best options through optimization. optimization tools, with a less significant
carbon footprint and reduced side effects
– Discover better catalysts – special chemical on plant growth.
compounds accelerating chemical reactions that
serve a defined purpose. Catalysts are key to: The applicability of quantum computers to model
the nitrogen fixation process has been explored in
– Decreasing energy costs and speed of theoretical proof-of-concept by ETH Zurich
green hydrogen production and Microsoft.27
Quantum computers that can run quantum circuits – Similarly, credit scoring at the point of sale,
are naturally positioned to work with complex which considers more diverse and uncommon
systems where data has inherent structure32 and sources, such as social media behaviour or
many variables. Structure that is pervasive across a other patterns, can become a reality due to
large dataspace is extremely costly to determine with the increased speed and complexity of scoring
classical computers and solving many of today’s hard models. Such an approach could increase
computational problems with this structure comes profitability of credit offers and allow to reach
down to finding the balance between the time spent unbanked population.
finding the answer and its usefulness for the end user.
– Better securities and derivatives valuation,
With many industry sectors relying increasingly incorporating more market factors to price risk
on data for planning and operations, quantum in close to real-time. The ultimate extension
computers will allow industries to potentially of this is portfolio risk assessment, analysis
augment their optimization and machine learning and optimization, allowing for future behaviour
processes to find new insights and make better and to be modelled and for portfolios to be
more precise decisions. optimized according to given parameters – by
comparing various combinations in a much
Financial, investment and insurance products more efficient way. One notable example is
One of the most straightforward applications of the better performing sustainable investment
aforementioned concepts is in the financial industry. portfolios, taking into account more ESG-
Banks, credit unions and trading companies are all related information, scoring and emissions
dealing with a massive amount of data, probabilities disclosures. Lastly, quantum computers are
and assumptions in their decision-making. also more efficient in running Monte Carlo
(probability) simulations.
– Crunch billions and trillions of financial
transactions and associated data on location, Some of the largest financial institutions have
time of day, merchant history, payment habits already carried out inquiries to investigate the
and more to identify unusual activity and flag most promising use cases for industry preparation
potential fraud. Quantum computers can (BBVA33), or are working on their quantum
improve pattern recognition in structured and applications with clear timelines to be the first to get
unstructured data sets, improving the quality a competitive advantage once quantum computers
and speed of fraud detection. are more developed (Goldman Sachs34).
Deutsche Bahn started its work on exploring the In 2019, in response to inevitable computing
potential of existing quantum computers to be used limitations, Airbus launched its Quantum Computing
regularly in real-world train scheduling (stage 1) Challenge,36 asking competing teams around the
and increasing the network capacity and use while world to address five critical design problems for
minimizing costs (stage 2).35 developing a new aircraft using quantum computing
methods. The winning team developed a quantum
New product design algorithm that proved quantum computing could
Quantum computers can process highly complex optimize an aircraft’s payload capability to maximize
design tasks with numerous variables, both revenue, optimize fuel burn and lower overall
simulating the behaviour of the material or operating costs.37
The types of problems that quantum computing is year-long competition to develop quantum-
expected to be best at solving (quantum simulation, resistant algorithms. Four of these have
optimization, quantum linear algebra and prime formed a basis of an upcoming post-quantum
factorization) underpin the inner workings of cryptography standard.39
many existing technologies. On the one hand, it is
expected to accelerate machine learning algorithms – The second approach is to use quantum
and provide pathways for unprecedented random number generators (QRNGs) to
opportunities in fundamental science, from physics secure communications instead of pseudo-
to chemistry to biology and advanced materials. random numbers generated with today’s
On the other hand, it carries a disruptive potential techniques. Several quantum computing start-
for the security of today’s digital communications ups are offering QRNG devices.
and, by extension, national security. Blockchain
solutions, including cryptocurrencies, are also under – Lastly, the quantum key distribution technique
threat, although good news exists for these areas. (QKD) exchanges encryption keys using
quantum communications. These experiments
Cybersecurity have been successful for networks of limited
Modern digital systems, products and services rely geographical size, and development continues.
on a set of mathematical problems to secure data
collection, storage and exchange. These problems None of the three methods is a silver bullet able to
are beyond the reach of classical computers, but provide guaranteed protection from break-ins in
quantum computing can solve them much faster the new quantum computing era. Malicious actors
– up to a matter of minutes and days38 – making constantly adapt, but using at least one of the
encryption breaking both feasible and worthwhile. stated approaches – or a combination of them – will
reduce the risk to a manageable level.
– Shor’s algorithm, developed in 1994,
spearheaded the field of quantum computing as More details about quantum threats, as well as
it was a tangible quantum algorithm that could available solutions and what firms can do to protect
have serious implications for society. If used on a themselves today, are covered in the newest white
general-purpose scalable quantum computer (that paper on Quantum Security.
does not exist yet), it would break the encryption
of today’s digital communications, compromising Artificial intelligence
existing public-key cryptographic algorithms. A primary application of quantum computing
is solving machine learning algorithms more
There’s a lot at stake when it comes to digital efficiently, thus turbocharging existing and new
encryption: artificial intelligence (AI) applications, e.g. quantum-
assisted models for generating data to enhance
– Data breaches, especially sensitive health or weather forecasting.40 In some applications, such
financial data as quantum neural networks, the entire learning
technique is transplanted into the quantum domain.
– Private and business digital communication,
including objects of critical infrastructure like Machine learning problems are relevant in multiple
power grids industries, including, for example, autonomous
driving, automated trading, speech and image
– The integrity of digital documents recognition, and predictive maintenance.
To fully exploit the potential of quantum in calculation speed or quality, while the rest of the
applications, real quantum computers with real problem is being addressed by a classical computer.
hardware to run quantum circuits are required.
Although it is possible to use classical computation While all components of a quantum computing
to simulate qubits and quantum interactions, this stack are important and can influence the quality
can only be done up to a certain size, after which and the speed of the calculation, the underlying
there isn’t enough computational power. hardware and quantum circuits to be run are
arguably the most critical components that must be
So what is needed for a quantum computer figured out.
to work?
Quantum computing relies on a hierarchy, or “stack”, – Quantum hardware (e.g. the qubit processing
of components to ensure ease of use, interoperability, unit, or QPU) is where the actual quantum
scalability and reproducibility. Together, this stack circuits are run.
of interrelated technologies – physical hardware,
control electronics, programming languages and – A quantum circuit is a sequence of quantum
algorithms to name a few – allows instructions that operations applied on multiple qubits.
form quantum circuits to reach a quantum processor Combinations of the results of quantum circuits
and perform necessary actions on qubits to produce advance a program to solve a computation.
a computational result. Quantum computers can
operate independently or in tandem with classical Many quantum computing hardware platforms
computers, either in parallel or in sequence. For require specialized environments to operate within,
instance, a quantum computer takes up part of including cryogenic cooling, ultra-high vacuum and
the computation, where it can provide advantage magnetic shielding.
High hardware requirements mean that most commercial systems are offered
“as-a-service”, often in collaboration with leading cloud providers. This means that
organizations will likely be buying computing time, not quantum computers.
Still, it is a good idea to create clarity over “what’s (and hence applications) can be run on today’s
in the box” or the cloud and whether it is the imperfect quantum machines; the most promising
right option for an organization. Knowing the applications of quantum with provable advantage
state of development and quantum hardware (e.g. Shor’s) will need much more developed and
types is essential from a quantum application scalable quantum hardware to operate.
perspective. Only parts of quantum algorithms
In 2022, there Competing quantum computing systems under stages, level 3 is the most technically advanced
continue to be development today rely on various quantum and higher-level systems satisfy all lower-level
technological physical interactions to define qubits – electrical, requirements).
improvements optical and magnetic. The hardware and
conditions needed to create and control qubits While two hardware platforms are ahead on the
across multiple
differ according to the quantum activity being development path – superconducting qubits and
hardware platforms
controlled, e.g. a photon of light or the spin of an trapped ions – other platforms and approaches
with technological electron. Each approach has different trade-offs could still be able to catch up at a later stage. In
development in terms of investment and benefits. Again, there the long term, it is likely that multiple hardware
roadmaps to help are parallels to classical computing, which evolved approaches will co-exist and find their niche in a
forecast paths. over multiple modes of physical infrastructure – new quantum computation ecosystem, based on
from vacuum tubes to magnetic tape to silicon their inherent advantages.
circuits on chips.
Important to note that no system has reached the
In 2022, there continue to be technological fourth milestone (level 4) today and cannot operate
improvements across multiple hardware platforms at scale. Roughly speaking, “at scale” could be
with technological development roadmaps to help understood as the ability to control 1 million qubits
forecast paths. in a computation, theoretically allowing for the
creation of a sufficient number of error-corrected
Despite the differences in approaches, the basic qubits to demonstrate quantum advantage in a
milestones in the development of a quantum real-life application. Today, our best-performing
computing system are the same: platforms have demonstrated capabilities in the
low hundreds.
1. The system must be able to create well-
characterized46 qubits. Building systems that follow the four steps of
development require universal control and are
2. The system must allow for the qubits to be capable of running quantum circuits. These
initialized, universally controlled and measurable “quantum circuit-model platforms” all have a
for calculations. universal gate set, which is a set of gates that
allow any arbitrary quantum circuit to be run, and
3. The system must be able to correct errors combinations of quantum circuits with classical
inherent in the physical hardware realization computations drive algorithms to fuel application
of the qubits. development. These are critical pre-requisites for
building a general-purpose quantum computer
4. The system must be able to do all of the that can be integrated as part of a computational
above at scale. workflow. As mentioned prior, quantum computers
will not exist on their own in vacuum but must be
Based on the above, all hardware platforms can be architected into the appropriate classical computer
classified according to the milestones achieved by infrastructure for ease of use and flexibility for the
mid-2022 (see Table 4, where level 1 is in its early end user.
Nitrogen-vacancy
(NV) centres
Advantages: potential
for on-chip integration,
stability (long coherence
times), modularity (optical
interconnects).
News articles often present the number of qubits as entagled or even fully connected (here the number
a metric to evaluate new achievements in quantum drops to 24 qubits), and thus their use may be
hardware. The number of qubits explains how limited.
scalable the hardware is, but it doesn’t explain
anything about the quality or speed of the qubits One of the balanced ways to look at the pure scale
and quantum operations. Numbers may fluctuate of the platforms (see Figure 9) is to count the largest
significantly depending on which qubits are taken number of connected qubits with universal control
into account. For instance, neutral atoms platform (to prepare a non-trivial quantum state). Again, every
has demostrated 289 coherent qubits, the largest metric will have its caveats, and should not be used
achievement so far; however, not all of them are as a standalone measure for platform evaluation.
—
Number of connected
qubits with universal control
Phototonic
simulators
NV
centres 9
Superconducting
qubits
14
Trapped ions
102 24
Neutral
atoms
Spin
systems
24
Source: World Economic Forum, Global Future Council on Quantum Computing
For quantum computers, computation speed is which data is available. The larger the calculations
measured in the number of circuits (or algorithms) per second, the better. For actual computations,
executed in a given time. Therefore, “calculations there is likely the need to run a large number of
per second” for QC cannot be compared to circuits to improve accuracy. IBM has proposed
classical computers’ similarly sounding benchmark. circuit layer operations per second (CLOPS)48 as a
Figure 10 shows a general order of magnitude potentially more stringent metric for benchmarking
in the rate of execution for trapped ions, speed. So far, CLOPS has only been measured for
superconducting and neutral atom systems, for multiple superconducting qubit platforms.
When looking at the quantum computer as a whole, a full stack, classical control
electronics is a speed limiting factor for most approaches in the current state. It
creates a bottleneck with a qubit measurement time, slowing the error-correction
cycle. Until control electronics improves, circuit speeds, demonstrated by different
platforms, does not translate into application speeds and can be misleading.
—
Calculations
per second
Trapped
ions 100
Superconducting
qubits
Neutral
atoms 1
1,000
The quality of quantum computation is measured Figure 11). Note that for general performance
by the accuracy of the results for running quantum measures, there is the quantum volume (QV),49
circuits – looking at the error rates. Different which gives a single number based off of the
quantum circuits of different complexity will performance of the system across a set of
obviously have varying error performance (see randomized circuits.
FIGURE 11 Quantum computing platforms quality comparison: two-qubit gate error rates
—
Error rate Neutral
atoms
(smaller is better)
Superconducting
Photonic qubits
simulators
2.5 % 0.1%
NV
centres
Spin systems
1%
1.5%
Trapped
ions 0.1%
10%
Such a holistic measure attempts to throw out errors, connectivity, crosstalk and other unintended
hardware bias, and considers several features of the noise to give an overall better idea of the usefulness
underlying hardware, from gate and measurement of a quantum computer.
Crosstalk: Occurs when an operation on one qubit Connectivity: Describes the connections
affects other qubits unintentionally. Crosstalk is a between qubits in the quantum architecture.
major source of noise in NISQ systems and is a These connections enable entanglement between
fundamental challenge for hardware design. different qubits.
—
Quantum
volume
Trapped
ions
Superconduting
qubits
512
9x9
4,096
12 x 12
Other ways of benchmarking the performance developed a benchmark suite50 that can measure
of a system is an “application-driven” approach: the progress towards the performance of quantum
A combination of speed and quality, looking at computing on four specific algorithms and use
performing a specific algorithm to track progress cases. As new algorithms are developed they can
towards useful quantum advantage. QED-C has be added to the suite.
Today all quantum applications can be divided into As of today, quantum applications in use have
two types depending on the quantum computing yet to demonstrate their advantage in a real-life
hardware requirements: fault-tolerant (requiring scenario, and are not yet being used on a
hardware above level 3) and near-term applications day-to-day basis to power end user solutions
(can be run on level 2 and 3 hardware). at scale. Researchers and businesses are,
however, already experimenting with near-term
1. Fault-tolerant: Applications with provable quantum applications to evaluate the potential of
advantage over existing computational methods the technology – for their industry and to inform
that will likely require a more developed quantum further research efforts into new algorithms
computer capable of quantum error correction. and applications.
2. Near-term: Applications that could be run on In summary, the jury is out as to whether a quantum
today’s noisy quantum computers, although it advantage can be reached on a non fault-tolerant
is not yet clear if or when these applications will quantum computer. What is certain is that error
provide any advantage over similar applications mitigation techniques with current noisy quantum
run by classical computers. Bridging these from computers allow the running of a handful of
what is likely achievable in the near term with applications, good enough for organizations to start
what is proven on large error-corrected quantum their inquiries and get prepared instead of waiting
computers is the subject of intense study and for general purpose (level 4) quantum hardware to
exploration throughout the industry. be fully developed.
From a global perspective, every continent already is, at the doctoral level. Finding qualified individuals
has start-ups in quantum technologies, and this with previous work experience in the world of
number continues to grow. This business boom business or engineering in an already scarce talent
would require an exponential growth of jobs in pool is proving increasingly difficult.
quantum technologies over the next two decades.51
More importantly, an effort must be made to
There is a need for skilled labour in the field of ensure that there are enough people with the right
quantum technologies. More than half of quantum skills to fill this explosion of jobs in the next twenty
companies are currently hiring. These companies years. The only way to educate the workforce of
struggle to find people with the right skills for new the future is by introducing quantum concepts at
positions in the emerging quantum job market. the primary and secondary education levels and
Some have referred to this as the “quantum skills creating more opportunities and programmes
shortage”. The fact that quantum technologies specifically for quantum engineering. Educating
An effort must
are still in their infancy means that most current the future workforce is a long process, but there
be made to ensure
jobs are highly technical, especially with academic are already several higher education programmes
that there are specializations and PhDs. In the past year, however, worldwide focusing on quantum engineering52 (see
enough people more diverse profiles, such as marketing and sales Figure 13). The quantum workforce includes a
with the right skills roles requiring prior work experience, have begun wide range of skilled labourers, such as quantum
to fill this explosion to appear, showing that the market is maturing. physicists, computer scientists, engineers,
of jobs in the next Simultaneously, the only people trained in the field technicians and people with business, sales and
twenty years.. of quantum technologies are highly academic, that policy backgrounds.
Netherlands
United Kingdom
1
4 Germany India
Ireland
4 1
Canada 1
Czechia
1 2
Sweden
United States
11
Australia
2
Spain
5
France Romania
3 1
Hungary
Switzerland
1 Italy
1
Source: QURECA, 2022
1
Globally, many individuals, both students and development of national quantum education
professionals, are looking for additional resources to programmes and addressing diversity issues.53
help them build their careers in the field of quantum
technologies. New companies specialized in There is also the dilemma of bringing quantum
quantum careers focus on supporting people seeking technologies closer to potential end users and
employment in quantum technologies, providing specific business sectors that are not yet part of the
them with professional guidance and placing quantum ecosystem. Fortunately, there is a very
them as ideal candidates in quantum companies. promising scenario in terms of educational resources
Opportunities to build and sustain a quantum-ready for both individuals and businesses at different
workforce are abundant, including public-private levels of specialization. Numerous global education
partnerships, industry-academia collaboration, initiatives include games, online courses and events.
It is never too Now, in 2022, is an ideal time to create policies developing regulations around emerging and critical
early to consider regarding quantum computing, as the technology technologies; many include quantum technologies.
and address any itself is approaching the sweet spot when it is
socioeconomic increasingly clear what technology will be able to These policies include issues around export
issues posed by a do, and what impact can be expected on different control that will likely be defined in a handful of
sectors of economy, while the actual applications national initiatives, and hopefully aligned through
new technology.
are currently limited, so any regulation will still be international agreements. Risk may lie in individual
proactive but informed.54 The lessons learned from AI countries acting to control specific hardware items
demonstrate that it is never too early to consider and relevant to building a quantum computer for both
address any socioeconomic issues posed by a new economic and national security reasons. Given the
technology. The World Economic Forum published a potential value of quantum computing to social
Quantum Computing Governance Principles report55 good and economic competitiveness, it can be
to help policy-makers and governments develop said that broad international agreements are the
policies and regulations on quantum computing. most ethical and fair process in moving forward.
Moreover, the supply chain for building a quantum
The types of policies and regulations needed in computer is likely to be global, as it is for most
quantum computing include those pertaining to high-technology industries. Although, given the
common good, accountability, inclusiveness, current geopolitical situation, political incentives
equitability, non-maleficence, accessibility and will likely localize supply chains and aim for
transparency.56 The threat of quantum computing strategic tech sovereignty via measures like foregin
to cybersecurity also requires policies to mitigate direct investment (FDI) screening.
this risk. Moreover, in a quickly progressing NISQ
era, security takes on a new and unique role as an There are currently no policies regarding quantum
enabler. In January in the US, an executive order computing and its energy usage. Introducing such
was passed that brought high-profile attention to policies should be considered to ensure future
the quantum cybersecurity ecosystem, presenting scalable technology that isn’t at the expense of
a roadmap to transition all federal agencies to enormous energy costs.58 Lessons can be learnt
post-quantum cryptography in 2035. This order from classical computing, which is currently
was shortly followed by a presidential memorandum struggling to decarbonize and is responsible for
regarding quantum security and quantum roughly 4% of global electricity consumption, and
ecosystem development.57 1.4% of global carbon emissions. There should
be no delay in promoting green quantum
Many countries, including those in Europe, North computing if technological choices are to be made
America, Asia and Australia, are increasingly during its development.
As quantum computing continues to mature, some Right now, the most valuable activities are
level of standardization across several aspects not so much around formal standards but the
of its development could help enable better accumulation of benchmarks for characterizing the
communication, benchmarking and compatibility hardware’s overall performance and creating a set
across various products – while mitigating risks of of standards that can be the basis for comparing
hindering technology development. For example, various quantum hardware.
standards and ecosystem agreement could be
effective in the following areas: Standards play a foundational role in market
development, especially for emerging technologies
– Terminology for better communication such as quantum computing. By defining common
terminology, hand-offs between companies,
interfaces between devices, and best practices and
– Benchmarks and metrics for evaluating
benchmarks, standards are the glue that holds a
quantum algorithms and quantum hardware
supply chain together.
– Standards for developing quantum software, Nevertheless, the areas for standardization should
algorithms and languages be chosen carefully. Due to the proliferation and
evolution of multiple hardware platforms for quantum
– Guidelines and best practices computing, it is too early to develop standards for
many aspects of this technology. For example,
– Certification and test protocols. photonic quantum computing will need optical
Some level of interconnect standards to connect photon sources
standardization to photonic processing chips to photon detectors.
across several Multiple international standards organizations have Since there are many different schemes under
become engaged and have begun to discuss the development using different wavelengths of light
aspects of
path forward for future standards. These include ITU and different ways to encode information, custom
[quantum
Telecommunication Standardization Sector (ITU-T), connections needed for one company’s system
computing] Institute of Electrical and Electronics Engineers will not necessarily be relevant for other technical
development could (IEEE), European Committee for Electrotechnical approaches. This problem is not limited to photonic
help enable better Standardization (CEN-CENELEC), International computing, since different ion species are at the
communication, Organization for Standardization (ISO), Corporation heart of different trapped ion computing platforms,
benchmarking and for Education Network Initiatives in California the components they connect to will need to work at
compatibility. (CENIC), among others. different wavelengths, making standards premature.
Type of standard/
Organization/document Description
pre-standard
ISO/International
Terminology Electrotechnical Commission This joint ISO/IEC committee is developing a standard specifically
for quantum (IEC) for terminology and vocabulary for quantum computing. As of March
computing 2022, it was in the committee draft stage.
JTC1 WG14/4879
Terminology for
A small team of experts is drafting proposed standard terminology
single photon
Under development by NIST for single photon sources and detectors, expected to be submitted
sources and
to a standards development organization in late 2022.
detectors
Standardization
CEN-CENELEC Focus Group Motivation to start developing European quantum technology
roadmap
on Quantum Technolgies standards in the pillars: quantum computation, quantum
for quantum
(FGQT) communication, and quantum metrology, sensing and imaging.
technologies
Use case
Use case for quantum technologies, complementary to the
for quantum CEN-CENELEC FGQT
roadmap.
technologies
Source: World Economic A key area that deserves focus is the development differences in underlying performance, changes
Forum, Global Future of application and performance benchmarks that in connectivity and eventual changes in error
Council on Quantum enable the comparison and evaluation of different correcting codes.
Computing
technological implications of quantum computing,
Acknowledgements
Achyut Chandra Naveen Kumar Malik
Lead, Open Innovations, HCL Technologies Director, Accelerator (Exploration and Incubation);
Chief of Staff at CTO Office, HCL Technologies
Antonio Córcoles
Research Staff Member, IBM Thomas J. Watson Justine Lacey
Research Center Research Director, Commonwealth Scientific and
Industrial Research Organisation
Jan Ole Ernst
PhD Candidate, University of Oxford Ulrich Mans
Strategic Partnerships Lead, Quantum Delta NL,
Mrityunjay Ghosh Netherlands
Quantum Computing Principal, HCL Technologies
Thomas Monz
Koen Groenland CEO and Founder, Alpine Quantum Technologies
Quantum Innovation Officer, QuSoft GmbH
Travis Humble
Deputy Director, Quantum Science Center, Institute
of Electrical and Electronics Engineers