Quantum Computing: Ready for the Huge Leap? Foreword

QUANTUM COMPUTING: READY FOR THE HUGE LEAP? FOREWORD

Quantum: the next frontier for startups?

What is going to be the next “buzzword” By increasing and accelerating comput- doing the rounds in our tech ecosystem? ing capacity, quantum computers have 2015 saw “big data”, 2016 was flooded the potential to transform many parts with mentions of “artificial intelligence”, of tomorrow’s economy—with banks, and in 2017 everybody was talking about pharmaceuticals, and transport the first “blockchain.” Each of these breaking to be affected. Cryptography, artificial waves shared a key feature: to draw the intelligence, and even the modeling of www.francedigitale.org attention of the general public and insti- life forms will experience a technological Founded in 2012, France Digitale is the largest startup association in Europe. tutional investors to major technological acceleration. France Digitale brings together the champions of digital entrepreneurship: it advances. Faced with the progress of American and gathers 1400 digital startups with strong growth plans and more than 100 investors (venture capitalists and business angels). Yet of all these tech waves, quantum com- Chinese giants, Europe needs a compre- puting might just prove to be the most The association (non-governmental organization) has a specific DNA, it associates hensive “quantum strategy” to engage entrepreneurs and investors to make the ecosystem more conducive to the powerful. At present, this breakthrough research institutes, scaleups, large emergence of new champions. technology remains a relative unknown, groups, investors, and public authorities. concentrated in the hands of several big Quantum computing is too important a tech players, despite its apparently mas- subject to be left just to engineers. It’s sive potential for application. The Positive Way time the life forces of the digital ecosystem seized it. This is the main mission of France Digitale, the leading association for startups in Europe. We hope this work www.wavestone.com will provide the foundations of a common Today, knowing how to adapt and drive transformation are the keys to success. At framework. Wavestone, we believe that a shared sense of enthusiasm is at the core of successful change. That’s what we call “The Positive Way”. “The Positive Way” is how we drive change in our clients, it’s who we are and it’s our commitment to creating a positive impact for all our stakeholders. Wavestone has more than 3,000 employees in 8 countries. It is one of the leading Nicolas BRIEN independent consulting firms in Europe and the leading independent consulting firm CEO France Digitale in France. Wavestone is listed on Euronext in Paris and is a Great Place To Work® listed company.

2 3 SUMMARY

03 Foreword

An ecosystem that’s rich... but still young Our convictions on quantum 24 06 Research institutes as key partners Digital giants in the vanguard Industry players: the vital catalysts Investment funds are essential allies What will the impact be on business sectors? 08 Countries as strategic players—and their funding programs The ecosystem in France and Europe: many projects... but few Quantum computing: ready for the huge leap? startups Quantum computers aren’t just more powerful versions of conventional computers! The simulation of complex systems The optimization of large systems Getting ready for quantum computing Cybersecurity 34 Artificial intelligence and quantum computing: a happy marriage? For public decision-makers In short: quantum computing will affect almost all sectors For the private sector

The fundamentals of quantum computing 18 Conclusion A brief history of quantum computing 38 The fundamental principles: qubits and superposition Measuring a quantum computer’s power The problem of technological standards Will quantum computing replace traditional computing? Our convictions 3 on quantum

Europe has every interest in developing quantum technology in order to remain in the race, maintain its leadership, and mitigate critical risks (such as the outsourcing of R&D to the US or China, falling behind on productivity, and the inability to protect sensitive information such as classified documents or economic intelligence). 1 4

Quantum computing will be a source of progress in all sectors both finance and industry, it will then be a catalyst for research discoveries. Europe needs a significant boost fromMember States if it’s to become a major global player: it needs to put in place a large-scale plan to stimulate investment, pursue training initiatives, and coordinate all its 2 stakeholders more effectively.

It poses a major risk to cybersecurity, threatening to 5 render current encryption systems ineffective. But it is also a source of solutions to overcome this risk. It is also a source of solutions for securing our communications. France now occupies a prominent place (2nd) in the European quantum ecosystem: French startups and manufacturers must transform quantum technology into uses and immediately seize its opportunities; investment funds have a key role to play in seeing the emerging technologies take off.

6 7 What will the impact be Quantum computing: on business ready for the huge leap? Google announcing its quantum supremacy… IBM marketing its first quantum computer: this time, the big players are making big moves!

Since the middle of the last century, the race has been on to increase compu- sectors? ting power. Computing problems are becoming increasingly complex: there are ever-greater volumes of data to exploit and organizations need to move faster to remain competitive.

Quantum computing promises to drive real progress in the area: Google claims it can reduce calculations that would take 10,000 years with a standard computer to a processing time of a few minutes.

But does it offer a miracle solution? And does quantum alone represent the future of computing? To better understand the coming revolution, and help you prepare for it effectively, in this document, we’ll shed light on the issues, quantum’s current and projected uses, its ecosystem, and the actions you need to take.

9 Quantum computers aren’t To summarize for now, the quantum prop- The simulation of complex / Physics just more powerful versions erties used enable the same processor to systems Research laboratories in calculate different problem “histories”, of conventional computers! By applying its powers to the infinitely France and Europe spent whereas a traditional computer calculates billions of hours using Before setting out on our analysis of this small, the quantum computer is an excel- them sequentially (or in parallel, but on supercomputers to solve complex upcoming revolution, it’s essential to be lent tool to simulate how matter and life multiple processors). equations—such as predictions about clear about the fundamental character- function, as things that play out at pre- particle collisions and nuclear reac- istic of a quantum computer: its ability The time savings can be considerable... cisely this scale; chemistry, physics, and tions. In the future, quantum comput- to process certain types of calculation in but only in certain cases: when the prob- biology, then, are prime areas to exploit ing will reduce calculation times and parallel and simultaneously. lem involves working through a large the quantum’s scope. number of possible combinations! enable work on reactions that remain Quantum computers represent an entirely The facilitation of discoveries in these undiscovered today. new model of computing, based on prin- Given this, there are three major types fields holds considerable promise—and For example, it should provide a better ciples quite far removed from their tra- of problem that will benefit fully from rich rewards for humanity and certain understanding of the keys to creating ditional cousins—just as light bulbs and advances in quantum computing: the sim- industries: superconductors that can operate at candles are only distant relations. In both ulation of complex systems, the optimiza- room temperature, something that cases, the overall objective is the same tion of large systems, and cybersecurity. / Chemistry - could drastically reduce losses from (to give light or perform calculations), but These problems are found in almost every Pharmaceuticals - national power grids, or increase the technologies used, and the associated sector of activity, something that merits Health a rapid overview of the use cases envis- the performance (and/or reduce the potential, have very little in common. The computational power aged, or even in use today. costs) of magnetic levitation trains needed to design new molecules (such as Japan’s Maglev, which is today is too large for current technol- scheduled to enter service in 2027). ogy to cope with: quantum computing will make it possible to simulate Quantum computing alone can’t make complex structures and offer a con- these applications a reality, but it will crete tool to design molecules with undoubtedly be an important catalyst. “The overall objective is the targeted properties (such as drugs, To date, the simulations possible with catalysts, proteins, enzymes, etc.) in a quantum computers have been limited same, but the technologies much more focused way than today. to very simple systems (comprising a few atoms); much more powerful quantum In a similar way, it should also improve computers will be needed to exploit the used, and the associated our understanding of parts of living full potential. organisms, such as the functioning of potential, have very little in the brain. common”

10 11 The optimization of large For example, Airbus launched its Cybersecurity becomes trivial: research to define Quantum Computing Challenge in such post-quantum cryptography systems So far, we’ve been mostly talking about early 2019, calling on the quantum is underway, but as ever, the transi- Leaving the atomic level, large systems, the opportunities offered by quantum community to help solve some of its tion from theory to practice, with an composed of thousands or even millions computing: the areas where it could strategic problems, the majority of acceptable level of standardization, of interacting parts, are also good candi- become a major lever of performance. which are optimization issues (wing will probably prove complicated. dates for quantum computing: its ability But quantum computing’s power is not shapes, aircraft loading, etc.). to simultaneously process all possible without its issues: cybersecurity, vital to /...but also as the “stories”, and therefore avoid traditional safeguarding the internet traffic that has / Finance ultimate solution! “combinatorial explosions” in these types become essential for modern societies, of areas, makes perfect sense—if the aim Financial institutions also provides the best illustration. But quantum physics also consume large amounts offers the definitive solu- is to rapidly identify optimal situations. This concept can be applied to a wide of computing power, / Quantum computing as tion to securing exchanges: the sen- range of areas: especially to assess the risk expo- a major cyber risk... sitivity of particles to any observation sure of their portfolios. This results makes it possible to detect, with abso- One of the primary appli- in hundreds of thousands of daily lute certainty, if information has been / Transport - Energy - cations of a quantum com- transactions. By drastically reducing intercepted on the path between its Logistics puter is the decryption of the keys the calculation time for risk levels, transmitter and its receiver. Logistics experts, such used to encrypt information: in doing quantum computing should further And the road to this solution may not as energy specialists and so, it generates risk for communica- improve the performance of the algo- be that long: China in particular has transport operators, face a daily chal- tions, payment systems, the validation rithms that drive a growing proportion made strong progress in the area, lenge to optimize numerous systems, of financial transactions, and all block- of today’s transactions. having put in place a ground-space, without the option of being able to chain applications—like bitcoin. The financial sector, which combines a quantum-communication link for its calculate the effects of all the pos- These keys are based on the fact that mature need and significant resources satellite, Micius, back in 2017. sibilities (at least in an acceptable factorizing very large numbers as to invest, will almost certainly be at time). Examples are route planning, prime numbers is difficult, a problem These technologies will undoubtedly the forefront of quantum computing. the management of fleets of vehi- that lies exactly within the core capa- constitute one of the security bricks of cles or aircraft, the arrangement of bilities of a quantum computer; Shor’s tomorrow’s internet and will strongly / Marketing containers in ports or on ships, the algorithm (1994) was specifically impact the telecommunications sector. optimization of supply chains, the Marketing departments face designed for this purpose. The beginnings of this are already vis- management of distribution networks complex challenges, such as ible, via, for example, the Quantum ID However, it’s not time to panic yet: (water, energy, etc.), and weather fore- optimizing their marketing solution, which, since early 2019, has this application won’t be feasible for casting. Quantum computing should mix, advertising placements, or even played a part in 5G-network security a good ten years given the complex- greatly improve and facilitate these predictive marketing. Quantum com- for a South Korean mobile phone ity of current keys. It is, however, time optimization calculations. puting could accelerate the resolution market leader (SK Telecom). to move to find solutions that could of all these problems. withstand this type of attack once it

12 13 Artificial intelligence and quantum computing: The impacts are sometimes less direct... but more quickly seen! a happy marriage? Quantum computing also means fundamentally revising the way algorithms are Artificial intelligence (IA) has been the star of the computing world in recent years, designed. This can be viewed negatively (because it forces us to learn new ways of thanks to the impressive nature of machine learning (or deep learning), but can it reach doing things) but sometimes there can be benefits too. a new milestone using quantum computing? For example, in 2018, Ewin Tang, a student at the University of Texas, improved a traditional machine learning algorithm to bring it up to the same level of performance A catalyst for acceleration as the quantum algorithm published in 2017. Compared with previous methods, this At first glance, and without a doubt,AI is hungry for computing power. It often resulted in an exponential change in the efficiency of the calculations that make movie, aims to solve optimization problems, making it an obvious candidate for quantum book, and meeting recommendations. computing. Without calling into question quantum’s potential, this anecdote highlights an inter- Thanks to its combinatorial power, quantum computers should enable reduced learning esting edge effect: current quantum algorithmic research can have positive short-term and processing times for many AI applications. Some examples are: benefits.

/ Finance - Public Sector A catalyst—certainly. But not a magic wand One of AI’s major applications today is detecting fraud and abnormal behaviors, a real Quantum computing is already driving the discovery of new algorithms and can con- weak spot for the financial sector and public-sector security departments: quantum tribute significantly to reducing the energy consumed in long-term AI learning. It computing should help to better and more quickly identify fraud-related patterns. will help to facilitate access to these technologies, and give a second wind to AI, by unlocking a new wave of development: intelligences based on quantum power. / Automotive - Mobility Yet quantum computing is not a panacea, and it won’t be able to address AI’s current A combination of AI—to make vehicles autonomous, and quantum computing—to key weaknesses: centrally optimize systems with thousands of parts, has huge potential here. / Explaining algorithms: the ability to explain AI decision-making is currently lim- Some significant initiatives have been widely publicized: in 2017,Volkswagen and ited—and a source of mistrust for users. D-Wave demonstrated the benefits of the technology to optimize the use of 10,000 / Algorithmic bias: the systematic errors made by algorithms as a result of the data taxis in Beijing, and, in 2018, Ford and NASA (again working with D-Wave) used it to on which they were trained; and, more generally, the sensitivity of the algorithms develop autonomous vehicles. to the data used to train them.

14 15 In short: quantum computing formed, resulting in large players making ing devices (using time, position, gravity, ers can reasonably expect the market to will affect almost all sectors major strategic shifts that will gear them underground soundings, etc.) that will be offer them solutions that will address this more toward organic growth based on capable of unparalleled precision. new risk. This brief overview demonstrates the con- in-house discoveries—and less focused siderable potential for progress across a But this overview also raises undeniable By now, we hope we’ve convinced you on the targeted acquisitions that have whole range of sectors: today, who can risks, which need to be carefully moni- that it’s worth understanding the subject become the norm in recent years. honestly say that they’re not facing any tored, especially when it comes to infor- in more depth... time to focus on quantum challenges related to resource optimiza- The “big tent” of quantum technology mation security issues—a topic high on physics, the research labs that produce tion? For some sectors, the impacts are also covers other innovations. To highlight the agenda for today’s organizations. the qubits, and the trails being blazed in even greater. To name but one, pharma- just one, think about quantum metrol- On this point, major private-sector play- the race to quantum supremacy. ceutical R&D could be completely trans- ogy, which promises to create measur- A brief history of quantum there was no need to give in to quantum computing effects: instead we should take advantage of them to move beyond the limits The fundamentals Moore’s law, which empirically predicted of conventional computers. a doubling of the computing processor power every 18 months, came into being From then on, research in the field has in 1965. Up to this point, it had proved been increasing across all areas: hardware true, mostly as a result of the efforts to of quantum components, algorithms, languages, etc. miniaturize the transistors used to build In recent years, we’ve seen an acceler- the processors. But, as a result of this ation in this technological revolution: need to miniaturize, by definition, physical quantum computers are being marketed limitations came into play: at very small and/or are accessible in the cloud (for computing scale (broadly the atomic level), quantum example, those of D-Wave, Rigetti, and effects can no longer be ignored, some- IBM); development environments are thing that completely disrupts operability. being used by communities (for exam- Even though we’re not there yet, the con- ple, LIQUi|> and Qiskit); and algorithms cept of the quantum computer originates are being optimized for quantum compu- from this observation, and especially the tation (for example, the factorization of genius of physicist, Richard Feynman. In large integers into prime numbers, Shor; 1981, he explained he was convinced that and the research algorithm, Grover).

The main milestones in quantum computing’s development

First basic model of a quantum Theoretical description of Development of a quantum First 5-Qbit NMR computer. An operating standard that allows • Arrival of the D-Wave 2000Q, with 2048-Qbits. Temporal Defense computer. “Traditional the functioning of a quantum algorithm to factorize large Technical University of Munich the control of a 12-Qbit machine. Systems is the first company to buy one. D-Wave Systems. computers can’t simulate the computer and quantum numbers. This exponentially Followed soon after by a Institute for Quantum Computing • IBM unveils its 17-Qbit quantum computer. development of quantum gates. outperforms the best traditional 7-Qbit machine. with input from the Institute for • Intel confirms the development of a 17-Qbit processor. systems in an efficient way.” D. Deutsch algorithms. Los Alamos National Theoretical Physics R. Feynman P. Shor Laboratory • Google announces the creation of a 72-Qbit-processor, Bristlecone.

1981 1985 1994 1998 2000 2001 2006 2011 2015 2017 2019

First experimental First demonstration of Shor’s D-Wave One, the Arrival of D-Wave’s • IBM unveils its first commercial demonstration of a quantum Algorithm (factorization of the first quantum 2X, rated at 1,152- quantum computer. algorithm by a 2-Qbit quantum number 15). computer available Qbits. IBM Q System One computer using Nuclear IBM’s Almaden research center to purchase. D-Wave Systems • Google announces that it has Magnetic Resonance (NMR). D-Wave Systems achieved quantum supremacy. Oxford University

19 The fundamental principles: with 10 on 1024 states, and over a mil- / Its topology, which describes how Two types of computers qubits and superposition lion states are possible with 20 qubits: qubits are linked together. enough to consider a vast degree of / The number and variety of available UNIVERSAL The basic building block: the qubit parallelization. operators (the “quantum gates”), QUANTUM QUANTUM ANNEALER As everyone knows, the basic building The exponential (in its proper sense) which offer unit operations (the equiv- COMPUTER block of a conventional computer is the capacity of a quantum computer is in play alent of “and,” “or,” or “exclusive or,” in bit: with a value of 0 or 1 (depending on here, and this is how we can explain, in traditional logic). Description With an ability to Limited to certain whether current is “passing” or not); by a few words and a necessarily simplified These last two characteristics, when carry out any type types of compu- intelligently combining enough bits any way, how a quantum computer calculates, cleverly combined, reduce the number of calculation, tation, and slower information can be represented. it can be used than a universal in parallel and simultaneously, the different of operations needed to perform a in all the areas quantum computer, But in a quantum computer, there’s a “histories” that we’ve discussed above. calculation. mentioned above it nevertheless works well for some fundamental change: information is now The technology still needs AI-related applica- stored in a qubit (a quantum bit), which Measuring a quantum tions (for example, has the astonishing property (a quantum computer’s power to progress autonomous vehicles) property, obviously) of being able to take Having said that, at present, the under- Reading press releases from the major the values 0 or 1 (at the time when we lying technologies are still maturing, and players, the measure of quantum com- The main Rigetti, IBM, IonQ, D-Wave measure it), with a certain probability. But so some of the intrinsic performance puting power seems obvious: it’s the players Alibaba, Google, as long as you don’t measure it, its value characteristics need to be considered; in Microsoft number of qubits that counts. Google is both 0 and 1: it exists in several “super- particular: was making announcements about posed” states. Scale (number 20-53 2,048 72 qubits in 2018, IBM about 53 qubits in / The error rate, which measures the of qubits in We’ll see later that a number of options 2019, while D-Wave is already marketing reliability of calculations: the lower it 2019) are being studied to physically create a computer rated at 2,048 qubits and is is, the greater the scope to link oper- qubits: atoms, electrons, and others. All promising 5,000 soon. ations together and therefore per- sort of things become possible when you form complex calculations. The use A key indicator: Quantum Volume switch to the scale of quantum particles! But qubits alone aren’t enough of error-correcting codes can com- pensate for errors, but at the cost of The reality is more complex, because the Announcements in the market are dif- A key property: the superposition of reduced performance (as in conven- headline number of qubits is actually a ficult to compare in the absence of all, states tional computing). poor representation of the power of a or some of, the measures above, which The real “magic” of quantum comput- quantum computer. That also depends / The coherence time, which defines makes it difficult to mark out the leaders ing comes into play when operations on: the period when the qubit remains in quantum technology. are performed on a qubit: because each “coherent” or, to put it more simply, / Whether its nature is “universal” or However, IBM has been promoting an operation is performed on the qubit’s “exploitable”: the longer this time not: some quantum computers are interesting measurement of power since (two) different states. is, the more operations, and there- designed to deal with certain types of 2017: Quantum Volume, which provides fore complex calculations, can be But it’s when we combine several qubits problems only, so it’s difficult to com- a single, aggregate measurement of the performed. into a system that this concept gets inter- pare them with universal quantum various characteristics mentioned. In esting: the number of these superim- machines (which are more complex to homage to Moore and his famous law, posed states then increases very quickly. develop but have a broader scope of IBM also aims to double the Quantum With 2 qubits we can operate on 4 states, application); Volume of its computer every year.

20 21 The problem of technological standards carry out complex calculations through Moreover, as we’ve seen, the power combining the power of traditional com- offered by quantum computers is only As with any maturing technology, there is the associated problem of standardization. puters with that of quantum computing— valuable for certain types of problem: it’s and reducing calculation times. useless, for example, to try to harness it for the wide world of transactional pro- In the summer of 2019, Accenture filed STANDARDIZATION LEVEL cessing (like booking train tickets or a patent related to directing calculations sending emails) or graphics processing toward the appropriate type of computer. Qubit types There is no consensus to date on how to physically create qubits. (like video games or TV). Similarly, Atos already offers a powerful Each player is focusing on the topology it considers the most powerful and hoping it won’t hit quantum algorithm emulator alongside Finally, one last, decisive point: quantum a glass ceiling. its supercomputer offer. This should not computers won’t replace all traditional The most promising technologies seem to be those based on CMOS qubits (Intel), superconduc- be confused with a quantum computer machines because conventional comput- tors (Rigetti), trapped ions (IonQ), or cooled atoms (PasQal); and, so far, none of the routes proper, but the emulator none the less ers will still be needed to program and has been abandoned. helps grow familiarity with the character- control their quantum cousins, as the istics of quantum programming. schematic below illustrates. Programming Each manufacturer is developing its own programming language, compilers, machine lan- languages guages, and development platforms. Understanding quantum computing While quantum algorithms are widely shared, thanks to the Quantum Zoo initiative, which lists the algorithms developed for quantum computing, developers are faced with multiple choices when it comes to implementing them—all of which are related to the decisions made by manufacturers. Quantum chipset

Actuators While there’s a debate among researchers Will quantum computing initialize and steer the Quantum gates about quantum hardware, the software calculation perform qubits operations replace traditional 3 discussion is taking place in a different (e.g. lazers, microwaves, 1 computing? Input 0 setting: here it’s the developer teams that data electrical impulses) H 2 will have to understand this raft of lan- Quantum computers are not about to Qubits guages and bring the applications to life. replace personal workstations. There are organized within a register, 1 they store handled data and two reasons: their large size and their fre- 4 At this stage, it’s hard to predict which enable the calculation framework will dominate, but don’t doubt quent need to be cooled to temperatures Traditional computer Measuring close to absolute zero (but who knows, 6 developers’ ability to master a range Work tool Measurement quantify and allow of languages (just as they can master, new types of qubits might ease this con- instructions interface 7 interpretation by a 5 transposes the quantum state today, a large number of conventional straint in the future). traditional computer into measurable data languages, provided they’re based on On the other hand, quantum computers (e.g. magnetometer) common principles). are likely to be integrated with supercom- Output puter manufacturers’ datacenters. We data anticipate that HPC (High Performance (Usually) refrigerated enclosure Computing) market players will gradu-

ally gear up to enable their customers to quantum computing” – “Understanding by schema inspired Source: O.Ezratty

22 23 Like most recent technologies, quantum Digital giants in the computing needs a rich and diverse eco- vanguard system: from research, through financ- Several major groups, with the US majors An ecosystem ing, to marketing—all players need to be at the forefront, have developed in-house active to bring about operational solu- programs dedicated to quantum technol- tions that will have an overall impact. ogy, in order to use it themselves, market that’s rich... Research institutes as key it, or fund research programs. Chief partners among these are Google, Microsoft, IBM and Intel. As the technology is still in its R&D phase, Since 2014, has been building its institutes and universities are key players Google own quantum computer via its Quantum in the quantum ecosystem. The majority but still young Artificial Intelligence Lab (QuAIL), which of them collaborate with other research units or private companies. is hosted by NASA. Its investment has enabled it to become one of quantum At European level, many research proj- computing’s most important players. ects involving international collaboration Google’s research team claims to have 1 are in progress, such as the OpenSuperQ reached quantum supremacy, the point project which brings together collab- where quantum computing overtakes orators from Germany, Spain, Sweden, the capabilities of conventional comput- Switzerland, and Finland. Collaboration ers. However, Google has not yet officially between institutions from at least three confirmed this, and Sycamore, the quan- countries is also a condition for access- tum computer it used, has limited real ing funding from the European Quantum applications. Flagship program2. Microsoft launched its efforts in 2005 On the French side, research is both by funding the Station Q research pro- ongoing and high quality. An example gram at the University of California, Santa is the Quantum Silicon research proj- Barbara. In 2017, the company launched ect in Grenoble, which brings together the Quantum Development Kit, a program researchers from three French institutes that used its programming language (Q#) (CEA-IRIG, CNRS-Institut Néel, and to develop quantum algorithms. The kit, CEA-Leti) to help develop silicon-based available in open source, has been down- quantum processors. loaded more than 100,000 times.

1- http://opensuperq.eu/partners 2- https://ec.europa.eu/digital-single-market/en/policies/quantum-technologies

25 IBM, which has traditionally been very Rigetti Computing is a Californian com- Industry players: for its work on optimizing Beijing taxi active in basic research, does not want pany that aims to develop an entire quan- the vital catalysts routes3, is working in concrete fashion to be left behind. It has developed an tum computer: manufacturing the chips, on quantum’s possibilities thanks to its Large industrial groups also play a role in-house research group focused on the and developing the architecture and algo- partnership with D-Wave. in quantum’s development. Many have end-to-end build of its own quantum rithms that will make them work. At this set up their own research teams or In energy, Total is working with Atos computer; and this has recently material- stage, it provides a cloud-based platform funded applied-quantum-research proj- (QLM) to, among other things, simulate ized in an announcement about the 53-bit to develop algorithms using 36 qubits ects working with research institutes or and understand particle behavior, as well system that became available in the cloud (although it’s not clear whether this is an startups. Some economic sectors, espe- as optimize the logistics of industrial in October 2019. The company aims to emulation or a 36-qubit machine). cially transport, aviation, telecommuni- tools. In 2018, EDF too launched a ded- commercialize its quantum computer in Chinese companies are also competing in cations and energy, are more aware of icated project. the next five years and position itself as the technology race. In 2015, the Chinese quantum’s potential than others. the ecosystem leader. IBM’s global hub The PASQuanS project, which aims to e-commerce giant, Alibaba, created the for these technologies is Montpellier, In 2015, Airbus commissioned a dedi- develop a quantum simulator (cover- Alibaba Quantum Computing Laboratory. France. cated team to work within its Defence & ing both hardware and software), and Its team is working on a number of quan- Space Unit. The aerospace giant is col- is one of the projects supported by the Intel’s focus is on the production of quantum projects, including, since March 2018, laborating with QC Ware (founded in the European Flagship initiative, ensures the tum computer chips. Its aim is to be able the manufacturing of 11-qubit processors, US in 2014), a quantum software startup. concrete application of the technology to mass produce quantum processors, the emulation of a cloud-based quantum The goal is to use quantum computing developed through a committee of end quickly and cheaply. computer, and the development of quan- for optimization, and quantum simula- users that includes Airbus, Bosch, EDF, tum algorithms. In 2018, another Chinese D-Wave and Rigetti, two North American tion to create ultra-durable materials. Siemens, and Total. giant, the search engine, Baidu, cre- startups, are key players in the quantum Volkswagen, already mentioned above, ated an institute dedicated to quantum technology market. computers. D-Wave is a Canadian company often 3- https://www.frontiersin.org/articles/10.3389/fict.2017.00029/full In France, Atos is the leading French seen as a pioneer in quantum computing, player in quantum computing devel- since it was the first to market annealing opment. Launched in 2016, the Atos quantum computers. It should be noted Quantum program is the largest indus- that this type of “simulated” computing trial program in Europe. In July 2018, only works for some, quite particular, the company released the Atos Quantum types of calculations, and therefore isn’t Learning Machine (QLM) emulator, the destined for the anticipated uses of a first large-scale, ready-to-use system, general quantum computer. capable of emulating up to 41 qubits on traditional Intel processors.

26 Investment funds are technologies. Their number is still low, Countries as strategic The European Union compared with artificial intelligence or essential allies players—and their funding In 1999, the European Union invested blockchain. programs €550m in the Future and Emerging To date, about 150 venture capital firms In France, Quantonation, the fund Technologies (FET) program. Following have invested in quantum startups. Quantum computers have become a stra- created by Charles Beigbeder and the “Quantum Manifesto” signed in May However, the number of investment tegic issue for sovereignty and security. Christophe Jurczak is an exception. 2016 by more than 3400 players from funds that specialize in quantum only is Many countries have put in place national Other French funds, such as the XAnge the academy and industry, the EU trans- very low: no more than half a dozen in the R&D strategies, with levels of state fund- fund, however, have recently invested in lated its desire to increase its invest- world; quantum technology is seen as a ing ranging from €140m to €1bn. quantum projects. ment in quantum technology into action. risky investment because of the technol- Looking worldwide Quantum Flagship, launched in 2018, is ogy’s complexity and the lack of visible The number of investment transactions, a 10-year, European Commission proj- returns on offer. and the total sums invested, have grown The US and China are the two countries strongly in recent years. In 2012, three that have invested the most4. In 2019, ect that invests €1bn to support 20 proj- Current investments are mainly made by startups had raised funds, amounting to the US government passed the National ects in its first phase, and mobilizes a specialized funds focusing on research or a total of US$34m. In 2018, 24 startups Quantum Initiative Act. This $1.3bn state European academic community of more those dedicated to one or more quantum raised US$128m. program aims to encourage both R&D than 5,000 researchers. and education in quantum technology. In 2013, the UK was the first European China is the country that has invested country to launch its national quantum Distribution of European investment funds (2019) the most in communications and quan- program with an investment of €370m tum cryptography. In 2017, the country over five years, which explains the many began the construction of its “National university programs and (relatively) large Laboratory for Quantum Innovation”, number of startups seen in Britain today. UK 30% which is expected to begin activity in France recently put quantum computing 2019, at an estimated cost of $10bn. Since on its agenda with a parliamentary mis- Germany 27% 2006, the total level of Chinese public sion on quantum technologies. Led by investment has amounted to almost deputy, Paula Forteza, it was launched €2bn. in April 2019; and should result in the France 17% executive announcing a national plan by the end of the year.

Finland 10%

Austria 7%

Ex aequo Norway, 3% Switzerland, Spain

Source: Wavestone 4 https://www.ida.org/-/media/feature/publications/a/as/assessment-of-the-future-economic- impact-of-quantum-information-science/p-8567.ashx

28 29 The ecosystem in France and the set of startups in the European Distribution of European startups (2019) Europe: many projects... but Quantum Flagship initiative) reflects few startups the current state of the market well: creating sufficiently powerful and UK 22% To monitor the European market’s stable quantum computers is still the France 18% development, Wavestone is launch- priority. This probably means that ing its Quantum Ecosystem Radar this opportunities in other areas (software Germany 15% year (see the graph below). Some initial development, for example) are still to observations: be seized. Netherlands 10% / At European level, the ecosystem European startups in this field are still Spain 9% has about 90 players. The UK is in first scarce and often recent: most were cre- place with 20 startups, and behind are ated in 2018 and 2019. Usually spin-offs Switzerland 6% France, with 16, and Germany, with 14. from research projects, and still largely / The number of projects or startups focused on R&D, most haven’t yet gained Poland 3% focused on hardware and component clarity on their business model (which is, Austria 3% development (which also includes of course, to be expected for a startup). Sweden 2%

Estonia 2%

Bulgarie 2%

“With 18% of startups and 17% Italy 2% of venture capital firms in the Finland 2% quantum computing ecosystem, Norway 1% France is positioned as leader in Europe*“ Czech Rep. 1% Denmark 1%

Belarus 1% NB: As the quantum market continues to evolve, we encourage the reader to adopt a certain degree of caution regarding the accuracy of these figures, which are mainly intended to provide an overview of the current ecosystem. Source: Wavestone

30 31 SECURITY 2019 Panorama of the European quantum ecosystem

COMPUTER Security

Hardware SERVICE

ALGO MACHINE LEARNING INTERFACE, ERROR CORRECTION Consulting

CAPITAL VENTURES Software

PHOTONIC CRYOGENIC SENSORS Investors

Components HUBS Wavestone figures from the following sources: Investment fund XAnge - xange.fr BPI QCB Conference (June 2019) EUROPEAN PROOJECTS Data base Dealroom Ezratty, O. (2019) - Understanding quantum computing Quantum computing report - https://quantumcomputingre- port.com Press kit Flagship (2018) - https://www.iqclock.eu/ uploads/1/2/0/6/120640577/press_dossier_final_ Hubs Flagship embargoed_1.pdf

32 33 A computer revolution is imminent: it’s dif- 2) Encourage collaboration at ficult to put an exact timescale on it, but European level most announcements suggest we’ll see “Quantum sovereignty”, like “digital Getting ready the first concrete results in the next three sovereignty” can and must be thought to ten years, depending on how optimistic through at European level—by strength- or skeptical their writers are. ening cooperation. Due to the cost and However things develop, there’s still time complexity of the research involved, most for quantum to prepare, and the following recommen- quantum computer projects are collab- dations should help both public and pri- orations between several public (and vate players take up the challenge of the often international) institutions. Although quantum revolution, and the many oppor- already prioritized in actions being taken computing tunities it will bring. by the European Commission, this collaboration could be further strengthened even For public decision-makers under the current framework. 1) Focus on European “quantum Without any contradiction, creating sovereignty” national bodies, with the responsibility for coordinating research projects, linking the Quantum technologies will be key in country’s different players together, and the defense and security sectors. And steering relevant strategies, could also the mastery of quantum computers and facilitate this transnational coordination. communications will be major resources in building secure infrastructures and being 3) Boost investment able to draw on unprecedented com- Quantum technologies are still heavily in puting power. This will mean getting the the R&D phase, and their development entire value chain right: materials, produc- requires significant funding. ing components, applications, etc. Beyond the funding of institutions The challenge is to avoid missing the engaged in basic research, there’s a boat on the next great technological case for launching public-sector calls for wave, something now leading many deci- quantum projects. In the short term, the sion-makers to lament US hegemony, defense and intelligence sector could be which is driven by the current strength of a major sponsor. Investments by major the GAFA companies. public or private players could also be encouraged through a dedicated taxation policy.

35 4) Develop skills For the private sector 3) Begin to experiment—and 4) Send out development teams as understand scouts To enable quantum technologies to 1) Evaluate the impact of quantum develop, new skills will have to be computing on strategic priorities There are several routes to supporting In parallel, encouraging development deployed: this requires the training of this type of analysis and the resulting teams to become familiar with quan- engineers, doctoral students, profes- As we briefly explained in the first part action plans: straightforward technology tum computing’s new paradigms seems sors, and researchers—something that of this Insight, the opportunities are vast, watching, training decision makers, or a a promising way forward: they’ll already will need to be organized and financed at but only in certain areas: in short, the sim- one-off partnership with a quantum com- be operational when the time comes, the right scale. ulation and optimization of large systems. munity player (a manufacturer, publisher, and can also help decision-makers iden- All facets need to be considered, because It’s up to the leaders and their strategists university, startup, VC firm, consulting tify where within the business quantum’s the creation of high-performance qubits to integrate this new potential into their firm, etc.)—to define or prototype a use power could be harnessed. case specific to the business’s needs. isn’t the only race taking place: beyond strategic thinking by asking questions There’s no need to plan to purchase and research in physics, research into software like: “which of my products and services Businesses can draw on all the classic install a quantum computer to do this: the also needs to be boosted to pave the way could benefit from this increased com- tactics used to engage with an ecosys- QCaaS (Quantum Computer as a Service) for the emergence of future developers puting power?” or “Could it be a source of tem: calls for projects, co-development marketing model, which uses cloud- and algorithms—both essential in harness- new offerings or significant competitive of solutions, financing teaching chairs, based resources, means progress can be ing this new source of power. advantage?” competitions (like Airbus’s challenge, made with limited initial investment. discussed above). Once the potential 5) Increase quantum’s visibility 2) Evaluate the scope to reduce costs has been confirmed, then a dedicated Finally, and more broadly, it’s essential to The mirror image to considering quan- team needs to be put in place, working in raise awareness among the public in gen- tum’s potential to help companies grow “commando” fashion to act quickly and eral, and public and private decision-mak- is the equally important one of its appli- precisely. ers in particular. All the more so because cation in reducing costs: “are there any even the use of the term “quantum” is cost-related areas we haven’t addressed often enough to provoke negative reac- due to lack of computing power, but tions, like “completely incomprehensible,” which could be tackled in a quantum “complex” “impossible to describe or era?” imagine),”. We’ve seen previously that many opti- This sheds real light on the need for the mization questions could benefit from quantum ecosystem to be clearly visible this revolution: for example, decisions on to both business and the general public. allocating resources, something close to National plans can be exceptional tools all business leaders’ hearts, might be illu- to bring these technologies to the fore, minated by new algorithms. demonstrate educational excellence, and position a country as a recognized leader in the field.

36 At the beginning of the 20th century, the advent of quantum physics turned models of the world upside down, at least Conclusion for some of the period’s greatest minds. But the subject, which was dry, complex, and often counterintuitive, remained obscure for mere mortals.

Yet we all use products, every day, that would not have been possible without this new science. And these have a considerable impact on our lives: semicon- Benoit DARDE, Partner Wavestone ductors (transistors, microprocessors, and therefore traditional computing); photocells (photocopiers, digital cameras, solar panels, etc.); lasers (the key to the digitization of audio and video, but also Today, the race to develop and per- extensively used in industry); etc. fect quantum technology is global; and Our firm conviction is thatquantum com- Europe, despite being in the running, may puting carries the seeds of a similar rev- struggle to stay the course, especially olution: by the huge leap it promises in faced with the colossal investments being addressing problems at the heart of our made by the US and China. But the stakes societies, it will open up vast pathways of are very high: they involve critical issues analysis and optimization that are closed such as European leadership in areas like today due to a lack of computing power. R&D, securing sensitive communications, Its potential applications are legion and or simply maintaining competitive levels affect all sectors: for example, the cre- of productivity. ation of new molecules that enable the Beyond the remaining technical chal- treatment of even more diseases; the lenges, which shouldn’t be underes- simulation of complex, multi-million-com- timated, but, none the less, will be ponent systems to manage smart cities gradually resolved in the years to come, effectively; or the arrival of impenetrable we observe two serious obstacles that systems to secure communications. need urgent attention.

39 The first is the challenge of growing European ecosystems that are competitive compared with their US or Chinese counterparts. The keys to this are well known: the creation of centers of excellence, the development of partnerships between basic research and industry, training people for tomorrow’s skills, unlocking financing, and a willingness to take risks. Here, public authorities have a vital role as a catalyst.

The second is the problem of a sluggish digital ecosystem, something probably linked to the technology’s maturity. Yet, as soon as this becomes accessible, and it doesn’t seem far off, Europe will need all the richness and creativity that its startups can offer to translate the technology into everyday uses.

But we already have all the cards in hand to make Europe a leader in quantum computing. So, let’s take courage; let’s seize the opportunity right now, experiment, and build the future, “quantum” world!

40 AUTHORS Acknowledgements

Interviews: Florian CARRIERE - Senior Manager Cyril Allouche, Atos Florian is a Senior Manager in the Digital and Emerging Technologies practice. For Marianne Billard, Parliamentary Employee 20 years he has been supporting IT Departments and Business Divisions in their Oscar Diez, European Commission major transformations related to the deployment of digital technologies. In recent years, he has led major projects on digital workplace and smart territories, mobili- Alexis Du Peloux, XAnge zing innovative technologies (platform, RPA, chatbots, AR/VR, etc.) using modern Olivier Ezratty, consultant and author, approaches (agile, UX design, lean startup). Paula Forteza, French Parliament Deputy [email protected] Jean-Christophe Gougeon, BPI Robin GAYRARD - Manager Olivier Hess, IBM Robin is a Manager in the Emerging Technologies practice, specializing in Cloud and Elham Kashefi, CNRS and VeriQloud Digital Workplace topics. A graduate of École Centrale Paris, he began his career in Pascal Maillot, European Commission Paris in Cloud transformations and IT organization at Wavestone. He then joined the Bernard Ourghanlian, Microsoft Lyon office, where he has been working for the past 3 years with the major players in the Auvergne-Rhône-Alpes region on their IT strategy. Georges-Olivier Reymond, Pasqal [email protected] Kristjan Sigurdson, Creative Destruction Lab Olivier Tonneau, Quantonation Christian Trefzger, European Commission Agathe RABAYROL - Policy Analyst France Digitale Agathe Rabayrol is Policy Analyst at France Digitale, in the Public Affairs division. Our thanks to Olivier Ezratty and Olivier Tonneau for their thorough proofreadings. She specializes in deep techs and data. She collaborated with transport.data.gouv. fr within the public policy incubator of Sciences Po for the opening of transport data at the local level, and is particularly interested in issues of inclusiveness and legitimacy of digital public policies. [email protected]

To find out more: Marianne TORDEUX - Director of Public and European Affairs IMT (2019) - “L’avantage quantique : enjeux industriels et de formation” Marianne is Director of Public and European Affairs at France Digitale. She CNRS (2019) - Les promesses de l’aube quantique represents the interests of France Digitale’s 1500 startups and 150 VCs Ezratty, O. (2019) - Comprendre l’informatique quantique. members, on all aspects (duch as tax, financing, platform workers status, IA, Blockchain, etc.). Marianne is also a tax lawyer and participated in the creation University of Maryland (2019) - Machine Learning meets quantum physics of the Blockchain reglementation in France. CEA (2018) - L’ordinateur quantique, graal du numérique [email protected] BCG (2018) - The Next Decade in Quantum Computing D-Wave (2018) - Introduction to Quantum Computing University of Innsbruck (2017) - Machine learning & artificial intelligence in the quantum domain CONTRIBUTORS UE (2016) – Quantum Manifesto

This publication was produced with the contribution of Jules Jaunay ([email protected]), consultant, and Mélodie Lauque ([email protected]), Wavestone Communication Department.

42 43