Q-campus Background study IN SUPPORT OF BUILDING A Q-CAMPUS - REALISING A QUANTUM ECOSYSTEM IN DELFT

Transcription

1 2018 Q-campus Background study IN SUPPORT OF BUILDING A Q-CAMPUS - REALISING A QUANTUM ECOSYSTEM IN DELFT

2 Reading guide 3 1. Sketch of Quantum Technology 4 Quantum Technology - a paradigm shift 4 Quantum Computing 7 Quantum Simulators 10 Quantum Communication and Quantum Internet 10 Quantum Software 11 Quantum Sensing and Metrology Industry Expectations 13 The Universal Quantum Computing Value Chain 13 Dynamics of the value chain for Universal Quantum Computers and Quatum Communication 14 Business models 16 Expectations 16 Investments 18 Applications of Quantum Technology in industry Dutch Quantum Ecosystem Overview 23 Knowledge Development 23 Talent 26 Financing 28 Networks 30 Leadership 32 Services and infrastructure 33 Demand Quantum Start-up Incubation 36 Start-ups in a campus environment 36 Start-up incubation 37 Start-up incubation at Q-campus 38 Sources Campus environment case studies 41 Q-campus Background study Background study 1

3 WaterCampus Leeuwarden 41 Leiden Bio Science Park 46 Chemelot 48 High Tech Campus Eindhoven 50 Sources 51 Appendices 52 Appendix A1: Qubit Roadmaps 52 Appendix A2: Progress in number of qubits 53 Appendix A3: Method used for selection within EU funding data 54 Appendix A4: Selected Dutch QT related master programmes 55 Q-campus Background study Background study 2

4 Reading guide This document is meant as a background study for the report Building a Q-Campus - Realising a Quantum ecosystem in Delft. It provides: necessary knowledge to interpret the progress in quantum technologies (1. Sketch of Quantum Technology), the industry expectations and the expected value chain of these technologies (2. Industry expectations), and an in depth analysis of the Dutch quantum technology ecosystem and its readiness for further investment (3. Dutch quantum ecosystem overview). These chapters make up the groundwork for the investment decisions and business case laid out in the Building a Q-Campus report. Additionally, an expert opinion is provided on start-up incubation with regards to quantum technology (4. Quantum start-up incubation). An overview of other campus environments in the Netherlands serves as the basis for the campus design in the final report and the relevant research has been added here (5. Campus environment case studies). Q-campus Background study Background study 3

5 Communication Computing Simulation Sensing / Metrology 1. Sketch of Quantum Technology Quantum Technology - a paradigm shift Since their invention, conventional computers store data in transistors that function as on/off switches called bits. A multitude of bits on a computer processor form a memory which can be accessed by programmes to perform a calculation. Computer processes have become increasingly more powerful as chip costs have decreased according to the famous Moore's law, which predicts that the cost and size per transistor halves roughly every 18 months. However, in recent years it appears Moore's law is becoming more difficult to achieve. The size of the transistor is getting so small that Moore's Law is likely to reach a limit 1. As the limits of bit-based processors come into view, a new potential to process information presents itself in quantum mechanics. Engineering / Control Software / Theory Education / Training Basic Science Quantum Technology Figure 1: Structure of the Quantum Technology field as presented in the Quantum Technology Flagship report by the European Commission High Level Steering Committee ( ) Exploiting the quantum mechanical principles of superposition and entanglement it is possible to build quantum bits (qubits). A qubit that can be in two states at the same time and the states of multiple bits can be manipulated simultaneously. This allows for an enormous growth in processing information compared to classical computers, as two qubits can be in four states and three qubits can be in eight states, leading to an exponentially large system. Taking advantage of these principles, a quantum computer would therefore be able to solve certain problems that would be impossible with classical computers. The ability to harness superposition and entanglement characteristics thus becomes the basis for the development of Quantum Technology. Within the field of Quantum Technology, four application areas were defined in the Quantum Technology Flagship project 2. Next to quantum computing, quantum communication and internet and quantum sensing techniques are important in the coming years. For all four areas, development of software and hardware are crucial ingredients. This factbook 1 Morgan Stanley (2017) Quantum computing weird science or the next computing revolution?, New York, Morgan Stanley Research. 2 High Level Steering Committee (2017) Quantum Technologies Flagship Final Report Q-campus Background study Background study 4

6 distinguishes between Quantum Computing and broader underlying Quantum Technology. Where the primary focus will be on Quantum Computing, it also explores themes of the consequences of other applications of Quantum Technology. European Quantum Technology Roadmap The Quantum Technology Flagship project defines goals within the four fields of Quantum Technology. In Figure 1 the goals are specified per field. This Flagship project represents a strategic investment aimed at enabling Europe to lead quantum technologies, building on its scientific research, on an established and growing interest from major industries, and on ecosystems of high-tech SMEs. The High-Level Steering Committee of the project has determined milestones which entail the quantum technology field containing every roadmap for the next 10 years 3. 3 High Level Steering Committee (2017) Quantum Technologies Flagship Final Report Q-campus Background study Background study 5

7 Quantum Computing 3 years: demonstrations will be shown of quantum processors with more than 50 qubits. The most important features to reach for platforms is quantum supremacy with an architecture where unit cells can be scaled and mass manufactured. Experimental devices will be ready for >50 qubits for quantum simulations. 6 years: logical qubits will outperform physical qubits and infrastructure of hundreds of qubits will be developed. The first field tests in data centres will be deployed. More algorithms and applications are developed. Quantum supremacy will be established in solving important problems in science. Demonstrations of quantum optimisations are ready. 10 years: demonstrations of fault tolerant implementations of relevant algorithm should arise in architectures of scale. Technology will be ready to deploy 100's of qubits which can be operated by users at data centres or other non-research parties. Quantum simulators beyond supercomputer capability are used to solve modelling problems in material science and AI. Quantum Simulation 3 years: experimental devices with certified quantum advantage on the scale of more than 50 (processor) or 500 (lattices) individual coupled quantum systems; 6 years: quantum advantage in solving important problems in science (e.g. quantum magnetism) and demonstration of quantum optimisation (e.g. via quantum annealing); 10 years: prototype quantum simulators solving problems beyond supercomputer capability, including in quantum chemistry, the design of new materials, and optimisation problems such as in the context of artificial intelligence. Quantum Communication 3 years: development and certification of Quantum Key Distribution devices and systems, addressing high-speed, high-trl, low deployment costs, novel protocols and applications for network operation, as well as the development of systems and protocols for quantum repeaters, quantum memories and long-distance communication; 6 years: cost-effective and scalable devices and systems for intercity and intra-city networks demonstrating end-user-inspired applications, as well as demonstration of scalable solutions for quantum networks connecting devices and systems, e.g. quantum sensors or processors; 10 years: development of autonomous metro-area, long distance (>1000km) and entanglement-based networks, a "quantum Internet", as well as protocols exploiting the novel properties that quantum communication offers. Quantum Sensing and Metrology 3 years: quantum sensors, imaging systems and quantum standards that employ single qubit coherence and outperform classical counterparts (resolution, stability) demonstrated in laboratory environment; 6 years: integrated quantum sensors, imaging systems and metrology standards at the prototype level, with first commercial products brought to the market, as well as laboratory demonstrations of entanglement enhanced technologies in sensing; 10 years: transition from prototypes to commercially available devices. Figure 1: EU roadmap goals for Quantum Technology (taken from High-Level Steering Committee (2017) Quantum Technologies Flagship Final Report, image: Birch) Q-campus Background study Background study 6

8 Investments in Million USD Qubits Quantum Computing The most promising and game-changing application of Quantum Technology is Quantum Computing. It encompasses a variety of different technologies in and of itself. The main challenge in building quantum computers is decoherence, the fact that quantum particles change state too fast to observe and use in calculations. 4 Below we will describe the two main developments of quantum computing. Task specific Quantum Computers In the field of quantum annealers, devices are built with qubits for solving specific problems by modelling the problem in such a way that it is equivalent to finding the lowest energy point in a landscape. Quantum annealers can be very powerful in solving this certain type of optimisation problems,. To date, D-wave Systems is the only company building commercial machines that may function as quantum annealers. These systems contain up to qubits. The company has sold the machine to Lockheed Martin and has partnered with Google and NASA to solve hard optimisation problems. 5 The increase of investments and number of qubits are displayed in Figure 2. D-wave Investment Rounds & Qubit results $200 $180 $160 $140 $120 $100 $80 $60 $40 $20 $ Cumulative previous investments Investment Qubits Figure 2: D-Wave Investments and Qubit results (source: Crunchbase, image: Birch) The Quantum Enhanced Optimization (QEO) Research Program by iarpa (The Intelligence Advanced Research Projects Activity of the US) also investigates quantum annealing and has started a trajectory to realise 100+ Qubits for annealing systems. 6 The QEO goal is to realise a basis of design for application-scale quantum annealers providing a factor speed-up over classical methods. The 4 M. Schlosshauer (2004) Decoherence, the measurement problem, and interpretations of quantum mechanics. Rev. Mod. Phys. 76, Thomson, C. (2014) The Revolutionary Quantum Computer That May Not Be Quantum At All, Wired, May 20th Quantum Computing Report (2017), Qubit Count Scorecard. accessed on Q-campus Background study Background study 7

9 program funds several public and private research initiatives to reach this goal and is led by the University of Southern California. 7 Universal Quantum Computers While task specific quantum computers only have the ability to solve specific problems, a universal quantum computer has the potential to compute a variety of complex problems. Therefore, this technology is in theory much more interesting to pursue. A universal quantum computer however, is harder to develop. Today the first quantum processors are operational and can be experimented with, but scalability and stability are still the main challenges in this field. Qubit platforms To build a universal quantum computer a number of different architectures are possible. These architectures determine how difficult it is to build the system and how many qubits can be interconnected. Research groups tend to focus on one type of architecture and large tech companies have already made their bets on certain types which they think will succeed, by allocating R&D money. The superposition state of a qubit is typically very sensitive to environmental changes. In general, the greatest barrier in developing architecture is overcoming this sensitivity. From a hardware point of view, these errors can possibly be reduced to a certain limit by solving different problems in for instance material properties, fabrication and the connection between quantum elements. From a software point of view, quantum error techniques together with fault-tolerant computing can potentially solve this (hardware) issue. In Figure 3, the most important technologies are described with their main characteristics, advantages and disadvantages Quantum supremacy Currently, an important goal is to achieve 'quantum supremacy'. Quantum supremacy is the point where quantum devices can perform a computational task that lies outside the capacities of classical computers. Together with the University of California, Google states that quantum supremacy is reached at around 50 logical qubits system 11. This counts for example for sampling problems. The EU defines quantum supremacy as a system of 100 robust qubits 12. It is key to note however, that although the number of qubits is important, the values above assume that the quality of the qubits is high, that is to say that the time the quantum information can be maintained is long and the noise on gate operations is low. 7 iarpa (2017), Quantum Enhanced Optimization (QEO), Research Programs, accessed on The full descriptions can be found in Appendix A2 9 Morgan Stanley (2017) Quantum computing weird science or the next computing revolution?, New York, Morgan Stanley Research. 10 Acín, A., Bloch, I., Buhrman, H., Calarco, T., Eichler, C., Eisert, J., & Kuhr, S. (2017). The European Quantum Technologies Roadmap. arxiv preprint arxiv: Neill, C., et al. (2017). A blueprint for demonstrating quantum supremacy with superconducting qubits. arxiv preprint arxiv: , 12 Acín, A., Bloch, I., Buhrman, H., Calarco, T., Eichler, C., Eisert, J.,... & Kuhr, S. (2017). The European Quantum Technologies Roadmap. arxiv preprint arxiv: Q-campus Background study Background study 8

10 The key point here is that supremacy is already reached with an amount of qualitatively good qubits which is relatively small when compared to classical computers with many millions of bits, and demonstration systems are already being pursued. We are approaching this point quickly, which is why the technology is gaining momentum and interest from governments and private parties. Superconducting qubits Trapped Ions Topological qubits Semiconductor spin qubits Working principle Resonance microwave circuits embedding a Josephson tunnel junction Electromagnetic field confining ions Specialised topological space for qubits using e.g. Majorana particles Nanoscale devices trapping electrons and using spin as qubit Current status Most mature, commercially available systems Medium scale systems have been realised Particles observed, large investments but first qubit is yet to be built Two-qubit algorithms realized Barriers Scalability due to noise restrictions Fabrication techniques for qubits Proof of concept working qubit Qubit uniformity Key actors* Research institutes, IBM, Intel, Google, QuTech, Rigetti, QDev Research Institutes, IonQ, Alpine QT, Quantum Factory Research Institutes, Microsoft, QuTech, Nokia Bell Labs Research institutes, Intel, QuTech * list not exhaustive Figure 3: Team analysis on the Qubit roadmaps, for more information see Appendix A1. The approaches in Figure 3 are considered to have the most potential and research groups receive funding to develop them further. Companies are either backing one or multiple of these technologies and it is generally expected that in the next decade applications for commercial systems will be viable with one of these approaches 13. Development of chips with an increasing number of physical qubits is an ongoing race between mostly large high-tech companies, including Google, Intel-QuTech collaboration, IBM and Microsoft and startups as Rigetti. The quality of qubits on the fabricated chips is not demonstrated in all cases. 13 Other technological roadmaps include Neutral Atoms,, Photonic Qubits and NV Diamnod Qubits. QuTech is active in the NV Diamond roadmap, using this technology for Quantum Networks. Q-campus Background study Background study 9

11 Quantum Simulators Simulating quantum mechanical systems is known to be a difficult computational problem. When attempting to solve these problems, it is a possibility to use a well-controlled quantum system to study a less-controlled or developed quantum system, known as quantum simulation. Quantum simulators can be regarded as analog quantum computers and permit the study of (novel) quantum systems that are difficult to study in the laboratory, and which are essentially impossible to solve on classical computers. A quantum simulator is designed to explore specific problems, for instance in quantum chemistry and bio-molecular physics, where there are important open questions related to energy transport in photosynthetic complexes and catalytic cycles of high biological interest. The development and design of quantum simulation algorithms still requires a big effort, not only to function for current quantum hardware systems but also to incorporate the current knowledge on classical simulation algorithms on quantum systems. Quantum Communication and Quantum Internet Quantum Communication Another major application of Quantum Technology is through quantum communication, basically the transfer of a quantum state from one place to another. Several applications of quantum communication are already known, for example, to secure communication, synchronize clocks, also perform secure delegated computations on quantum servers in the cloud. Currently, the security of classical communication is provided by encryption through classical computers. Some widely used encryption types could be broken by future Quantum Computers. Quantum communication is provably secure even if the attacker has a quantum computer now or in the future - a feat that has been mathematically proven to be impossible for any classical encryption scheme. Quantum Internet If two or more quantum computers are optically connected to each other, a quantum computer network can be formed.by being connected to this network, qubits can be exchanged by the quantum network nodes. Like classical networks, a quantum network features different elements: End nodes quantum computers connected to the network on which applications are run. The demands for such quantum computers to accomplish useful tasks is thereby very modest compared to quantum computing algorithms: many known applications such as for example secure communication only require quantum computers capable of preparing and measuring one single qubit. The reason why already one qubit per network node allows tasks that are impossible to accomplish on a classical network is due to the fact that quantum network protocols gain their power from quantum entanglement for which two qubits one at each end point is already sufficient. In contrast, a quantum computer needs more qubits than can be simulated on a classical one in order to do something useful. Q-campus Background study Background study 10

12 Repeaters the task of a quantum repeater is so transmit qubits over long distances. Since quantum information cannot be cloned, conventional repeaters cannot be used and quantum repeaters use genuinely new technology. Network technology such as for example low loss switches to maximize use of existing network infrastrucutre. These are not inherently different than classical optical switches, but less losses are desirable to achieve a better performance due to the fact that qubits are often communicated using single photons which are easily lost. Networks at short distances form an avenue to scaling quantum computers: As the number of qubits of one computing system is limited at this point, choosing a modular approach can help researchers and engineers scale up quantum computing systems. When one or more quantum computers forming the network are geographically apart, one can speak of Quantum Internet. A quantum-computing cloud that is accessible through the Quantum Internet is an ultimate combination that would provide the possibility to perform secure quantum computing in the cloud. 14 Quantum Internet is envisioned to be used in for example 15 : Secure communication with the help of quantum key distribution Clock synchronization Combining distant telescopes to form one much more powerful telescope Advantages for classic problems in distributed systems such as achieving consensus and agreement about data distributed in the cloud Sending exponentially fewer qubits than classical bits to solve some distributed computing problems Secure access to a powerful quantum computer using only very simple desktop quantum devices Combining small quantum computers to form a quantum data center One of the big challenges to Quantum Internet is making long distance communication possible. At this point, long distance communication is hindered by signal loss and decoherence (loss of quantum coherence) caused by the transport medium of for instance an optical fiber. For classical signals, an amplifier can be used to enhance signal, but as qubits cannot be copied, this is not possible in quantum communication. The technology that is being developed to overcome this challenge is called a quantum repeater and is often based on the principle of quantum teleportation. Quantum Software Crucial to the operation and success of quantum computers is not only that the engineering challenges are overcome but also that specialised software is developed and tested that functions on a quantum computer. Quantum software requires an entirely new approach compared to conventional computers. New protocols, algorithms and applications will have to be developed in order to exploit the power of future quantum computers and global quantum networks Q-campus Background study Background study 11

13 Next to application software both the development of the quantum computer and quantum networks require research into new methods to efficiently control and operate such systems. Quantum Cryptography Generally speaking, in cryptology, a cryptosystem (also cryptographic algorithm) combines a message with additional information, a key, to produce a cryptogram. If this cryptogram cannot be unlocked without the key, it is secure. When this key is the same for both encryption and decryption, the system is referred to as a symmetrical (secret-key) cryptosystem. When the key for two communicating users are different, it is called asymmetrical (public-key). Quantum cryptography is the science using quantum mechanical properties to perform cryptographic tasks. Quantum Key Distribution Quantum Key Distribution (QKD) is an example of quantum cryptography that could help symmetric cryptography. QKD is used to produce the secret key for both communication partners and not to send a message, this can be done through a standard communication channel, e.g. an optical fiber. An important property of QKD is that the two communicating users can detect a possible third eavesdropping partner as this third partner disturbs the quantum system when measuring it, in this case, the key will not be produced. Thus, the security is guaranteed by physical principles, not mathematical complexity. Post-quantum cryptography Post-quantum cryptography is the science of cryptographic algorithms that should be secure against intrusion by a quantum computer, not necessarily using quantum mechanical properties. This is very important as popular encryption techniques as for example ECC and RSA (public-key techniques), used to secure modern communication, can be broken by using Shor s algorithm on a powerful quantum computer, once it would exist. Fortunately, there are also multiple cryptosystems that are believed to resist Shor s algorithm and thus part of post-quantum cryptography such as Hash-based cryptography, Code-based cryptography and Lattice-based cryptography. Quantum Sensing and Metrology Quantum Sensing and Metrology is a technique that either uses a quantum system, quantum coherence or quantum entanglement to measure a physical quantity. To certain extent, quantum sensing exploits the weakness of quantum systems: quantum sensors are highly sensitive to their environment. What this sensitivity is party relies on its reaction to external parameters. For a spin qubit, this could for example be its response to an external magnetic field. Furthermore, the sensitivity relies on the intrinsic sensitivity of the quantum sensor. The quantum sensor should preferably respond strongly to desired signals but not to unwanted noise signals, a tricky dispute. 16 Examples of quantum sensing systems that are used and investigated are neutral atoms, trapped ions, Rydberg atoms, atomic clocks, solid-state spins and superconducting circuits. 16 Quantum Sensing, C.L. Degen, F. Reinhard and P. Cappellaro. Rev. Mod. Phys. 89, Q-campus Background study Background study 12

14 2. Industry Expectations The Universal Quantum Computing Value Chain Current developers of Quantum Computing Technology follow a model that reflects that of regular supercomputers. The hardware in development is expected to provide supercomputer-like devices that can be used as a service by others to do calculations. This requires software to operate the quantum computer and run applications on it. Thus, a potential value chain for quantum computing will have a similar structure to the value chain for conventional supercomputing. Even so, it is possible that more novel applications of quantum computers are still to be discovered that will alter the structure of this value chain. This is congruent with the outlook on the use of Quantum Computing and Communication that is shared among most experts. Quantum Computing is expected to become relevant in the next 10 years, within a niche market for specific problems. This means that Quantum Computers gain a status equivalent to supercomputers, with a low number of devices used as a service by corporations. Software developed for these computers will likely be tailored to solving industry-specific challenges. Quantum Communication offers usable technology in a number of phases: point-to-point and short distance technology to perform quantum key distribution is already commercially available now. A next step in the next five years can be metropolitan networks in which several clients are connected to a central hub in a star-shaped network. This allows anyone connected to the hub to run elementary quantum network applications such as quantum key distribution for secure communication or password identification. Such star shaped networks can maximize the use of existing infrastructure of standard telecom fibers for quantum communication. In parallel, further development of end nodes can bring more complex quantum network applications within reach at short distances. Larger scale networks at pan European distances are expected to become feasible after 10 years. This is a current assessment based on the activities of developers and the proposed architecture of Quantum Computers. It may be subject to change as the proposition of Quantum Computers matures. Hardware Equipment and component manufacturing Qubit memory and processor System architecture and protocols Operating QC mainframe or datacentre Software operating system applications Figure 4: Stylised QC Value Chain (Team analysis based on Rodney van Meter and Clare Horsman: A blueprint for Building a Quantum Computer, October 2013; Cody Jones et al, Layered architecture for quantum computing, September Image: Birch) Q-campus Background study Background study 13

15 Each of the activities has specific challenges outlined below. The most likely expected scenario is that future computing will be a hybrid between quantum and conventional computing and this value chain will run parallel to, as well as overlap, with the value chain for conventional computing. Similarly, for quantum communications a value chain akin to the conventional internet is anticipated involving both hardware and software development, infrastructure providers as well as online service providers. Challenges: Qubit memory and processor: the development of technical building blocks for the quantum computer, that is a functioning two level qubit, storage and gate technologies and interconnection technologies. Quantum repeater: the development of a quantum repeater to transmit qubits over long distances Quantum computer and network architecture: the design of the interconnection and communication between blocks in order to enable fast processing, and routing, and a setup for error correction. This requires not only hardware but also software developments in error correction, fault-tolerant computing and protocols for the new architecture. Quantum programming: tools for compiling and methods for debugging programs, verification and testing of Quantum computed results. Application algorithms and protocols algorithms and protocols that leverage some essential feature of quantum computation and communication such as quantum superposition or quantum entanglement. 17 Software and algorithms that enable quantum internet, perform simulation of quantum systems, perform machine learning and can be used in cryptography. Functioning quantum software is also a prerequisite for the production and operation of hardware and architecture and thus permeates the entire value chain. Dynamics of the value chain for Universal Quantum Computers and Quatum Communication The following provides an overview of the dynamics for each of the links in the value chain and an overview of private organisations currently active in it. Key players (not exhaustive) Trends Risks & Opportunities Equipment and components Specialised suppliers in cryogenic devices and other supporting engineering BlueFors Cryogenics Oxford Instruments Montana Instruments Suppliers are focused on better solutions to the noise problem, which results in a competitive advantage. Supply chain is not yet developed or standardised, every device poses enormous engineering challenges. 17 McKinsey commissioned by Ministry of Economic Affairs (2015), Global development of Quantum Technology Market. Internal document Q-campus Background study Background study 14

16 Agilent Technologies Leiden Cryogenics Delft Circuits Rohde & Schwarz Zurich Instruments Anyon SingleQuantum Keysight Q-devil Hardware manufacturing Qubits Architecture Capital intensive organisations Google IBM Intel Microsoft Rigetti NTT Pushing for vertical integration with architecture and software. Hard technology: defining a new arena and racing for a dominant design, the capabilities do not exist yet. Possible winner takes all dynamic. A lack of standards, incompatible hardware, constant change, and uncertain roadmaps. Scaling of architecture will be challenging and can cause a shift in production locations. Mainframe/data centre Alcatel-Lucent Bell Labs MagiQ IDQ Nokia SK Telecom SurfSara Pushing towards first deployed demonstrator systems Software development Operating system applications Mostly small-scale startups and tech giants active in hardware development 1Qbit Q X Branch Cambridge Quantum Computing Qbitlogic QCWare Rigetti Microsoft (Q# language) Google (Quantum A.I. research & cloud) IBM (QISKit) Zapata Qu & Co Experiments charting the capabilities of quantum software, in operation of a QC, algorithms to run on a QC and accessing a QC through the cloud. Large players (hardware manufacturers) will secure position with standard setting and IP. Challenge: support of one platform versus adaptive to multiple platforms Room for diverse set of players (industry specific small players or software giants with specialised division) Table 1: Team Analysis based on statements by the Ministry of Economic Affairs, Rigetti, Quantum Computing Report et al. Q-campus Background study Background study 15

17 Business models Based on this value chain and industry analysis there are several business models imaginable that make use of quantum computing 18. Platform Supplier: will provide a quantum computer, access to it over the cloud, software to program it and other tools or training. Clients are thus able to run programs of their own design. IBM and D-Wave are currently using this business model and other manufacturers of quantum computers envision this as their future business model. Vertical Market Specialist: will provide training tailored to a specific area, combining quantum knowledge with industry expertise. Example today is Q X Branch which works predominantly for finance and insurance markets. Component Supplier: will supply the necessary equipment and tools for building Quantum Computers. There is an infrastructure of companies that are taking this component approach. This includes dilution refrigerator suppliers like Bluefors, suppliers of certain specialized waveform generators, photonics, and measurement devices like Zurich Instruments, and software providers like QCWare. Some developers of quantum computing processing chips may also want to take this approach rather than building the full machine. Quantum Under the Hood: will use a quantum computer as a means to offer a service for an end application. For example, a company may offer a cloud-based machine learning or computational chemistry, drug discovery, or logistics optimization service that uses quantum computing. This requires development of tailored algorithms. Expectations The value added in each segment of the value chain is expected to be greater the more the technology advances into software and applications, similar to classical computing. The building of quantum hardware will provide engineering challenges with opportunities for high-tech companies (similar to for instance ASML or NXP for classical computing chips). In designing quantum system architecture, software and algorithms, the diversity of possible applications may be large and there is more room for both start-ups and established organisations to foray into designing products and services that use a quantum computer as part of the business model. 18 Based on the analysis by the Quantum Computing Report (2017). How to Make Money in Quantum Four Basic Business Models, December 14th 2017, accessed on Q-campus Background study Background study 16

18 Quantum software & algorithms Quantum system architecture Quantum computer and device manufacturing Qubit assembly and integration Figure 5: stylised representation of value added differences between value chain activities Another expectation that is a consequence of the development of the quantum computer is the multitude of spinoffs based on quantum technology that will find applications in other industries. Developing quantum components and devices for the quantum computer will also lead to advances in other areas. A roadmap for the expected spinoffs has been drawn up by the UK Quantum Technologies programme 19. This roadmap is independent from the EU Flagship roadmaps years 0-5 years Components and equipment for research Quantum key point-point links Metropolitan quantum networks 5-10 years Quantum components for industry Quantum sensors Quantum processor networks at short distances Imaging First intractable problems solved through QC years Off the shelf Q- sensors Pan European Quantum Networks Quantum crypto Quantum based navigation Quantum key satellite communication Cloud based QC Quantum computer coprocessor in small scale computers. Figure 6: expected spinoffs in quantum technologies, based on the UK roadmap, not exhaustive, image: Birch The collective value of these spinoffs is estimated to rise to between 280 million and 2.8 billion for the UK alone, which gives an indication of the possible worldwide value. 19 UK National Quantum Technologies Programme, 2015, A roadmap for quantum technologies in the UK, London, Innovate UK and the Engineering and Physical Sciences Research Council on behalf of the Quantum Technologies Strategic Advisory Board. Q-campus Background study Background study 17

19 Market size in Million pounds Yearly market size quantum technologies (for the UK) years 5-10 years years years Prediction timescale Figure 7: yearly market size estimates for quantum technologies based on the UK roadmap. Image: Birch. Investments According to The Economist, there is an estimated yearly budget of $1,5 billion in public funding dedicated to quantum computing in the coming years, of which $550 million is spent in the European Union and $27 million in the Netherlands 20. In 2018 China announced a $10 billion National Laboratory for Quantum Information Sciences. Aside from public funding, quantum computing has attracted significant investments from the private sector, both in the form of venture capital and more conventional company investment from large tech players. Several large investments by corporate technology companies have been made public. Examples are: IBM has incorporated quantum computing in a large investment program for 5 years, which takes a part of $3 billion in research investments made in post-silicon microchips (2014). Investment by Intel in Qutech for compound superconducting and semiconductor qubits: $50 million (2015). Alibaba invests $15 billion to set up 7 quantum computing labs worldwide, following up on $23 million from Large private companies such as Google (Alphabet) and Microsoft keep investments in Quantum Computing undisclosed in order to protect their competitive advantage. However, an investigation of numerous start-ups in quantum technology reveals that interest in the field is exponentially rising. Several recent funding examples are: Trapped ion qubit producer IonQ has recently raised $22 million Estimates are from 2016 over A. van der Steen, RVO letter, (2015) Alibaba Group invests in joint quantum computing laboratory html viewed Q-campus Background study Background study 18

20 Quantum computer developer Rigetti obtained $64 million in a second round of investments (2017) 23. Quantum cryptography developer ID Quantique announced a strategic investment plan by SK Telecom of $65 million (2018) 24. The size and behaviour of the companies involved suggest that financial requirements are not the bottleneck in the development. They are able to provide the financial means and because of the high stakes there is a reasonable chance that any investments necessary for the development will be covered, regardless of the price. Through these investments, tech companies appear to be competing in a rat race to be the first to acquire a certain number of qubits or display quantum supremacy. For a larger overview of companies with significant investment in Quantum Computing see the table in Appendix A2. Surveying ~60 start-ups in Quantum Technology, we found 18 companies that had (partially) disclosed funding information, as seen in Figure 8. In 2017, Quantum start-ups attracted a total of $179 million in disclosed funding, see Figure 9. It gives an indication of the rising interest in different fields of Quantum Technology and leads us to believe that the actual investments are higher than what is being disclosed. The latest data on 2018 (august) suggests that this will continue, with already almost $27 million invested (currently published), almost entirely in software and communication start-ups using Quantum Technology viewed viewed Q-campus Background study Background study 19

21 start-up funding Figure 8: Quantum Technology start-ups disclosed funding (source: Crunchbase). Quantum annealment Trapped Ion QC Quantum communication $ Full-stack quantum computing Quantum software Superconducting QC $ $ $ $ Figure 9: start-up funding divided by company activity (source: Crunchbase) Q-campus Background study Background study 20

22 Notable is the 571% increase in disclosed funding for quantum start-ups in 2017 compared to Notable also is the increase in funding for quantum software companies and the investments in startups aiming to build their own quantum computer, in direct competition with the projects of IBM, Microsoft and Google. It seems, despite a lack of published investment details, the private sector is willing and able to do sizeable investments in quantum computing, despite the early stage nature of the development. Applications of Quantum Technology in industry The buzz for quantum technology is largely a consequence of the promise in both performing radically new tasks and optimising time and resource consuming tasks in existing industries. Although predictions for these industries rely entirely on the scientific advances made in quantum technology in the coming decade, some industries have started experimenting and have reached promising results. Below we sketch expectations for the impact of quantum computing and communication in various industries. New applications of Quantum Computing and Communication Industry Quantum activity Impact Prospect Chemistry & Pharma logical qubit quantum computers should be able to analyse or simulate (parts of) molecules quantum mechanical properties 25. This can be used for drug development, catalyst discovery and material design, a good example for application would be fertiliser reaction calculations (Haber Bosch process). Reduction in the costs of molecule development, helping e.g. to reduce the on average $ 2 billion cost of R&D of new drugs 26. Accenture and Biogen with the help of 1Qbit verified that a quantum enabled method for molecular comparison was as good or better than existing methods 27. Analysts expect that this technology is readily available before Some functional algorithms have already been designed and tested in simulations. Chemical & pharmaceutical companies have a large incentive to adopt computing power. Material science Aircraft, buildings, cars and other complex structures could be designed with the aid of quantum computers, simulating the properties of materials at a molecular level. Drastic reduction in computation time modelling complex structures in comparison to regular supercomputers. Airbus assesses that it may be able to shorten modelling time of Some of these problems are already being tackled with quantum annealers and simulators. Analysts expect that this technology is readily available before logical qubits requires a multitude of physical qubits in a quantum computer. viewed on viewed on viewed on McKinsey commissioned by Ministry of Economic Affairs (2015), Global development of Quantum Technology Market. Q-campus Background study Background study 21

23 the aerodynamics of a wing of a new aircraft from seven years to several weeks Aerospace companies have a large incentive to invest in terms of reduction in aircraft development time. Cyber security & Defence Quantum Key Distribution (QKD) allows for nonhackable communication. With a quantum computer, current communication becomes vulnerable, whereas QKD allows 100% safe data transfer. Governments gain access to quantum computers for surveillance, intelligence, communication and national security applications. Quantum Key Encryption must be used to shield communication from access by a third party with a quantum computer. First code break by IBM has been proven. The algorithms to do large scale prime factorisation have been designed but require estimated thousands of qubits to run. Within 10 years technological hurdles should be cleared to create a quantum communication network 31. Communication & internet A Quantum Communication Network allows for long distance communication. It is expected this requires quantum repeaters that amplify a quantum signal beyond distances of 200 km 32 Quantum computers can connect through quantum internet to solve larger problems. Exponential savings in the amount of communication required to solve certain tasks are expected. QuTech aims to build a demonstrator in 2020 that connects four Dutch cities into a quantum internet. Within 10 years technological hurdles should be cleared. Machine Learning and AI Several cross-industry applications for QC's involve the increasing demand for computational power, whilst classical computers are becoming less cost effective in recent times to supply that demand. This is most noticeable in industries where data is a big part of the business, and large amounts of data analysis are done in real time to complement the core activity. This includes sectors such as the transport branch, which uses large amounts of data to build predictive models for traffic flows, the oil & gas branch where companies have live data on thousands of oil wells and any other branch involved in big data computations and optimisations viewed on High Level Steering Committee (2017) Quantum Technologies Flagship Final Report 31 High Level Steering Committee (2017) Quantum Technologies Flagship Final Report 32 Centre for Quantum Computation & Communication Technology (20170, Australia, viewed on Morgan Stanley (2017) Quantum computing weird science or the next computing revolution?, New York, Morgan Stanley Research Q-campus Background study Background study 22

24 3. Dutch Quantum Ecosystem Overview Whether the Netherlands is well suited to lead the developments in Quantum Technology through entrepreneurship and economic activity is dependent on the quality of the ecosystem that enables it. We distinguish six crucial elements in this ecosystem. Each of these elements will be described to determine the position of the Netherlands. Knowledge Development Talent Financing Networks Leadership Infrastructure and services Figure 10: Conditions for a successful ecosystem, based on Stam & Spigel (2015). Finally, we consider the Dutch industrial position in relation to Quantum Technology, to assess whether industries in the Netherlands are ready to adopt Quantum Technologies and whether the Netherlands can attract investors and companies to the Netherlands in Quantum Technology. Knowledge Development Knowledge development is defined as the opportunity for scientific advancement and the possibilities this shapes for knowledge transfer. Knowledge development of Quantum Technology and connected enabling technologies as photonics, material science, nanotechnology, data science takes place at several research locations. The Netherlands has three specialised institutes: QuTech, QuSoft and QT/e, and six universities active in quantum technology: the Delft University of Technology (TU Delft), the University of Amsterdam (UvA), the University of Groningen (UG), the University of Leiden (UL), the Radboud Universiteit Nijmegen (RU) and the University of Twente (UT) where research is performed. Compared to other European countries, the Netherlands has numerous quantum technology research locations across universities within very close proximity to each other, each with their own specialisation. Together, they cover almost all aspects of the current quantum technology landscape, and many of these institutes are at the forefront of their field of specialisation. Dutch quantum technology institutes QuTech (Delft) QuSoft (Amsterdam) QT/e (Eindhoven) Established in 2013; National Icon since Partners TU Delft (Faculty of Applied Physics and Faculty of Electrical Engineering, Centre for Mathematics & Informatics UvA VU TU Eindhoven Q-campus Background study Background study 23

25 Mathematics and Computer Science) TNO Mission develops quantum technologies based on superposition and entanglement aimed at scalable quantum networks and quantum computers. 34 Focus areas Fault-tolerant Quantum Computing Quantum Internet and Networked Computing Topological Quantum Computing Quantum Software and Theory to develop new protocols, algorithms and applications that can be run on small and medium-sized prototypes of a quantum computer. 35 Four research lines Quantum simulation and few-qubit applications Quantum information science Cryptography in a quantum world Quantum algorithms and complexity This research has the potential to revolutionize a wide range of fields, including ICT and the internet, simulation and computation, data science and security, energy and sensing technologies, healthcare, logistics, and materials sciences. Research topics: Post-Quantum Cryptography Physics-based Quantum Security and Quantum Networks Quantum simulators Quantum Nanophotonics Quantum Materials & Devices Funding TU Delft TNO Industry funding Public funding (EZ, HTSM TKI, NWO/FOM and STW). Funding from UvA s research priority area Quantum Matter and Quantum Information (QM&QI), a joint effort between four research institutes of the Faculty of Science. The center is currently hosted and funded by CWI, located at Amsterdam Science Park. TU Eindhoven FTE / personnel 2018: 200 fte 2023 (goal): 350 fte 2018: 60 fte 2022 (goal): 120 fte 2018: 10 PI s, 20 PhD s, 10 postdocs 34 (viewed on ) 35 (viewed on ) Q-campus Background study Background study 24

26 University quantum technology research throughout the Netherlands The following overview is not exhaustive. Group (G) or Department (D) Mission Known personnel UvA Quantum Matter and Quantum Information (D) Five UvA research groups are active in the area of Quantum Matter & Quantum Information, bundling the expertise of three UvA research institutes This research priority area is focused on the experimental and theoretical study of Quantum Matter (QM) and applications in Quantum Information (QI). QM & QI works on multi-particle entanglement, topological protection, quantum cryptography and theory of strongly interacting quantum matter. ± 18 Pi s TUD Quantum Nanoscience Department (Kavli Institute of Nanoscience) (D) RUG Quantum Devices (G) UL Quantum Matter and Optics (D): UT Quantum transport in matter (G) The Quantum Nanoscience Department, from which QuTech was launched in 2013, focusses on exploring new quantum devices and technologies using state-of-the-art nanoscience. In addition to its strong link to QuTech, the department currently has concentrated efforts in the fields of quantum materials, quantum sensing and mechanical quantum technologies. The RUG focus is on fundamental studies of quantum coherent dynamics in solidstate devices. The department has three topics, each with dedicated PI s: Quantum Information Quantum Matter Quantum Optics The research addresses quantum aspects of electronic transport in novel materials and devices. Examples of materials under study are correlated electron systems such as novel superconductors, oxides interfaces, and topological insulators. State-of-the-art materials science and nanotechnology is combined with ultrasensitive transport measurements to reveal novel quasiparticles such as Majorana fermions and magnetic monopoles PI s ± 60 PhD s ± 20 Postdocs 1 PI 4 PhD s 12 PI s 1 PI 36 (viewed on ) Q-campus Background study Background study 25

27 RU Quantum Matter (D) The goal within this research theme is to understand, develop, and manipulate materials based on collective, or emergent quantum effects, envisioned toward new types of functionality 20 PI s involved in programme, not all fulltime. The recent EU QuantERA call for Quantum research proposals demonstrates the research qualities of these institutes and groups. In a highly competitive call beginning with 91 full proposals, only 26 were awarded funding, of which 9 contained a Dutch partner (see table). Project Description Dutch Partner QCDA QUANTOX QuaSeRT Si QuBus SuperTop Quantum Code Design and Architecture QUANtum Technologies with 2D-OXides Optomechanical quantum sensors at room temperature Long-range quantum bus for electron spin qubits in silicon Topologically protected states in double nanowire superconductor hybrids Delft University of Technology Delft University of Technology Delft University of Technology Delft University of Technology Delft University of Technology ORQUID ORganic QUantum Integrated Devices Leiden University QuantAlgo Quantum algorithms and applications Stichting Centrum voor Wiskunde en Informatica QuompleX Quantum Information Processing with Complex Media University of Twente NanoSpin Spin-based nanolytics Turning today s quantum technology research frontier into tomorrow s diagnostic devices Wageningen University Talent Talent describes the present pool of (future) researchers and entrepreneurs in Quantum Technology that have the opportunity to contribute to the ecosystem. In Figure 11 (left), the number of MSc graduates in studies of direct interest to Quantum Technology. For the selection method, see Appendix A4. The number is displayed for the years In general, an increasing trend from 2012 to 2016 can be observed in the number of MSc graduates (+27%). In Figure 11 (right) the information on the universities where these MSc graduates (in this 37 data from DUO (2018), Q-campus Background study Background study 26

28 case in 2016) is added. This overall increasing trend in Quantum Technology related degrees in the Netherlands suggest a suitable talent pool for ecosystem growth Total number of MSc graduates in QT related studies per year MSc graduates in QT related studies in Male Female Figure 11: Left: Nationwide number of MSc graduates over the years in in studies related to Quantum Technology. Right: Distribution of MSc graduates over Dutch universities. Aggregated from DUO data. See Appendix A4 for selection method. Quantum technology education programmes In the QuTech Academy, students from Applied Physics, Electrical Engineering, Computer Science, Mathematics and Embedded Systems are invited to enrol in the Master programme of QuTech Academy. This Master programme consists of MSc courses (1 year) and MSc projects (1 year) providing the students with an excellent education to start their career as a PhD student in QuTech. Furthermore, online courses of top scientists from QuTech are offered via QuTech Academy. In Amsterdam (QuSoft) and Leiden modules are created for education on Quantum Computation, at MSc, BSc and outreach level. The NWO Zwaartekracht program Quantum Software Consortium will coordinate and strengthen education and outreach at QuTech, QuSoft and the University of Leiden. Q-campus Background study Background study 27

29 Financing Financing comprises the funding opportunities that research and commercialisation efforts in Quantum Technology have gotten in the past and may get in the future. The recently formed Dutch Government has formulated Quantum Computing as one of the top funding priorities but has not yet disclosed the amount of funding it will dedicate to this subject. In 2015, the Ministry of Economic Affairs committed to financing QuTech in a national partnership, together with several partners for 135 million over ten years, until All the partners contribute either in kind or in cash to this budget. This budget has been increased to 145,3 million for 10 years until 2025 by the PPP supplement scheme of the Ministry topping 25% on the private contributions. Within this budget, The Netherlands Organisation for Scientific Research (NWO) finances two industrial partnership programmes in QuTech with Microsoft, both on the subject of Topological quantum computation and Majorana qubits. In total, NWO finances 20,2 million between 2011 and NWO has also awarded an 18.8 million grant from the Gravitation Programme to the Quantum Software Consortium, a collaboration between scientists from Delft, Leiden and Amsterdam. There is a dedicated consortium in the Netherlands for nanotechnology, NanoNextNL, with a 250 million programme from 2010 until The consortium consists of more than one hundred companies, universities, knowledge institutes and university medical centres. One of the roadmaps is Beyond Moore, providing funding for nano devices that are relevant to the development of Quantum Computers. 39 In an international context, the Netherlands, with the Technical University of Delft primarily, is one of the largest receivers of European Scientific Funding in the field of Quantum Technology. Based on published data 40 from the 7 th Framework Programme, the European Research Council (ERC) and the Horizon 2020 Programme, it is estimated that the Netherlands has received more than 123 million research funding in 93 projects related to Quantum Technology between 2008 and 2018 (with projects starting in 2018 extending to 2022), seen in Figure 10. The total selection comprises 411 projects with almost 550 million in funding from FP7 & Horizon 2020, and an additional 43 Dutch projects with 89 million in funding from the ERC. The method for selection is explained in appendix A3. In the European context the Netherlands has the highest funding intensity per project and organisation and takes in 8,6% of European funding from FP7 & Horizon the partners are the Ministry of Education, Culture and Research, the University of Delft, The Netherlands Organisation for applied scientific research TNO, The Netherlands Organisation for Scientific Research and the topsector High Tech Systems & Materials. See Partnerconvenant QuTech, retrieved at retrieved on Data downloaded from retrieved on , and from retrieved on Q-campus Background study Background study 28

30 EU Funding in M Number of projects EU Funding in M Number of participated projects EU Funding Projects Figure 12: EU Quantum funding distributed across countries per project participant (countries with > 10 million Euros in funding) Within the Dutch Quantum Technology related projects, the TU Delft has received 73 million in 44 projects over the past 10 years, as can be seen in Figure 11. Here, ERC data has been added to gain The TU Delft attracts the largest amount of funding in the field of Quantum Technology within the Netherlands and Europe. Important to note that other universities and research institutes in the Netherlands also contribute to or coordinate 49 other Quantum Technology related projects with sizable contributions from the EU. The University of Amsterdam is the second biggest attractor of funding with 13 projects and over 15 million in EU funding Total funding Number of projects Figure 13: EU funding of Quantum Technology related projects and number of projects receiving this funding for different Dutch universities. Q-campus Background study Background study 29

31 Networks Networks are described in both scientific networks and science-industry relationships that benefit the growth of the ecosystem through distribution of knowledge, labour and capital. The EU research programmes selected in the previous paragraph are also an indicator of network formation, as almost all projects consist of large research consortia that perform joint research. Using the data on the participants within EU research programmes, it is possible to chart collaborations between universities and other organisations to reveal (a part of) the scientific network. Figure 14 is a graphical representation of this network, where nodes represent universities and other organisations (colour coded by country) and the links represent collaboration in a EU research project. The size of nodes represents the amount of funding they receive. Connectedness is measured in the number of collaborations (links) and number of unique partners (connection to other nodes). Q-campus Background study Background study 30

32 ULeiden UvA TU/e CWI Germany TU Delft United Kingdom France Italy Spain Radboud Other The Netherlands Figure 14: Network diagram of European Research Projects, organisations are scaled towards the FP7 & H2020 collaborative funding they receive. Between 2008 and 2018, the TU Delft participates in 103 collaborations in Quantum Technology with 67 unique organisations. The university has built an international European research network in which it takes up a central position. Within 8 of its projects it is the project coordinator, and within 10 projects the university also functions as the host institute. Despite their lower budget, other Dutch universities such as the TU Eindhoven and University of Amsterdam also take up central positions in the network through intensive collaboration. Q-campus Background study Background study 31

33 Leadership Leadership is formulated as a measure of guidance in the direction of Quantum Technologies research and commercialisation, in which public parties play a role in facilitating efforts. Dutch National Icon ( Nationaal Icoon ) QuTech Dutch National Icons are initiatives that deal with societal challenges using technological solutions and promise economic growth in the future. In 2014, QuTech was rewarded with this status (one of four organisations in the Netherlands) because of the promising quantum technological research performed there. As a National Icon, QuTech receives extra support and funding from the Dutch government. QuTech, together with a broad spectrum of other research institutes and private partners is part of the larger Holland High Tech Roadmap on nanotechnology 41 that coordinates application driven technological development in research institutes and private companies. National Science Agenda: Nationale Wetenschapsagenda (NWA) The NWA was defined with the idea that fundamental science in the Netherlands should focus on societal challenges now more than ever. Within the NWA, taking the connection to the European science agenda Horizon2020 into account, twenty-five routes in which the Netherlands can take a leading role were identified. Of these routes, the route Quantum-/nanorevolutie was one of the eight routes rewarded with 2.5 million euro s. Within this route, three main topics ( game changers ) are defined: Quantum Computing and the Quantum Internet, Green ICT and Nanomedicine. Coalition agreement 2017 The most recent coalition agreement Vertrouwen in de toekomst (Confidence in the future) published by the Dutch government explicitly notes that technical sciences with high research expenses will receive extra attention and that priority is given to the fundamental research done in line with the National Science Agenda. Furthermore, the agreement specifically states that policy in top sectors (dedicated to collaborations between industry, knowledge institutions and government) will be strongly focused on three societal themes of which one is defined as quantum/hightech/nano/photonics. Quantum Software Consortium In 2017, the Quantum Software Consortium (QSC) was rewarded 18.8 million euros from the Gravitation program financed by the Dutch Ministry of Education, Culture and Science. With this funding, a group of excellent researchers from disciplines computer science, math and physics can perform top research for a period of ten years to explore, develop and demonstrate possible applications for quantum computers. The group is a collaboration of researchers from the CWI, QuTech, the TUD, the UL, the UvA, the VU and is led by QuSoft. 41 Holland High Tech (2015), ROADMAP NANOTECHNOLOGY, viewed on Q-campus Background study Background study 32

34 Services and infrastructure Services and infrastructure comprise organisations and materials that are required for successful research and start-up support. The services capabilities of Delft are centred around YES!Delft. YES!Delft is a non-profit incubator created in It is funded jointly by the municipality of Delft, TNO and the TU Delft, but also receives corporate funding. In 2013 a second building dubbed YES!Delft 2 was added with 5500 m 2 of office and lab space. It focuses on high-tech start-ups and works mostly in the energy, clean tech, medical & health, industrial solutions and mobility clusters (YES!Delft has not (yet) declared quantum technology as a focus area). Its services cover a range from renting office and lab space to offering coaching programs. With its two main offers to entrepreneurs being: LaunchLab: 10-week (part-time) pressure cooker programme to validate a technology Incubation Programme: full-time growth programme where you build and grow your technology company. During the first six months, entrepreneurs take part in an intensive programme to work on all the basics of your company. YES!Delft has been the top-ranking incubator in the Netherlands in the last few years, according to UBI. Furthermore, it ranked 4 th and 2 nd in the category University-linked Business Incubators & Accelerators in Europe, in 2016/2017 and 2017/2018 respectively. With regards to quantum computing, both Delft Circuits and Single Quantum are part of YES!Delft. NanoLabNL NanoLabNL is a national organisation that facilitates nanotechnology research in 5 different cities in the Netherlands (Amsterdam, Delft, Eindhoven, Groningen, Twente). Facilities and expertise of NanoLabNL can be used by universities, research institutes, start-ups and industry. Nanotechnological research activities are crucial to quantum technology research, for instance in fabrication of quantum devices with high precision. NanoLabNL is funded by Dutch governmental programmes as NanoNed and NanoNextNL. Demand Crucial for the future uptake of a quantum technologies industry in the Netherlands is the ability of both local industries and global corporates established in the Netherlands to exploit the discoveries that are made working towards a quantum computer and other quantum technologies and create demand for these products and services. It is expected that the scientific progress in quantum technologies will lead to spinoff technologies that will lead to the creation of new devices and services (see Industry Expectations). Dutch industry position A good way of identifying a country s strong industries is reviewing export data. The Netherlands characterises itself in having a strong competitive advantage in the production of highly specialised machines and instruments, with companies that are specialised suppliers with very specific and complex technical know-how. Of the net exports of the Netherlands in 2016, 13% concerns machinery Q-campus Background study Background study 33

35 of which the details are given in Figure 15. Within machinery, 44% comprises specialised machines and instruments, with a total net value of almost 6 billion US$ 42. On average, the Netherlands has a high revealed competitive advantage in these products, exporting more than twice the expected value of specialised machines and instruments 43. For some particular sectors, such as equipment for photographic laboratories and medical devices the Netherlands has a far higher competitive advantage than the average country. Figure 15: Export of machines and instruments from the Netherlands in 2016, divided by subcategories and percentages of the total value, source: Atlas of Economic Complexity These industries are likely to be affected by quantum (spinoff) technologies in the next two decades and judging by the current industry positioning the Netherlands has the industrial expertise to adopt these technologies in highly specialised and complex manufacturing processes. A second expectation is that the value-added in quantum computing will largely come not from hardware development, integration and manufacturing but from software development that is successful in designing applicable programs and algorithms for a quantum computer (most likely used as a service). This means that fostering demand for the application of quantum computers requires a strong communication infrastructure and software-based industry. 42 Data from The Atlas of Economic Complexity, viewed on The Atlas of Economic Complexity defines a revealed competitive advantage as A measure of whether a country is an exporter of a product, based on the relative advantage or disadvantage a country has in the export of a certain good. ( ) a country is an effective exporter of a product if it exports more than its fair share, or a share that is at least equal to the share of total world trade that the product represents (RCA greater than 1). For more information see Q-campus Background study Background study 34

36 Export value (in US$ billions) The Netherlands is a strong ICT country, with over 30,000 specialised companies in software, internet and communication and in % of all service exports are related to this sector (see Figure 16), with a total value of 101 billion US$. Export value of IT services from the Netherlands as part of total service exports Other Service exports (BoP, current US$) Computer and communications services (BoP, current US$) $180 $160 $140 $120 $100 $80 $60 $40 $20 $ Figure 16: Export value of ICT services from the Netherlands. Data: World Bank, Image: Birch Furthermore, the Netherlands has an excellent infrastructure, with almost triple (2,903) the secure internet servers per 1 million people than the EU average (996) and a high rate of broadband dispersion (42.2%). A good example of large scale internet infrastructure is AMS-IX, the Amsterdam Internet Exchange, which is a neutral and independent Internet Exchange based in Amsterdam, the Netherlands. It interconnects around 800 networks by offering professional IP exchange services. Q-campus Background study Background study 35

37 4. Quantum Start-up Incubation By Chris Eveleens Start-ups in a campus environment Start-ups are innovative young businesses that have the aim to ambitiously grow. These types of firms are known to contribute to introducing innovations to society, creating jobs, and economic prosperity. Increasingly, initiating and developing start-ups is seen as something that occurs not in isolation, but in interaction with the context of the start-up (Aernoudt 2004; Autio et al. 2014; Acs et al. 2017). Therefore, there is an interest by both academics as well as practitioners in the reciprocal relationship between start-ups and their environment. The system perspective prescribes that not only the success and focus of start-ups is contingent on its context, but also vice versa that over time the start-up affects its context. In a virtuous cycle, the context supports the emergence and growth of start-ups, who then strengthen their context by feeding resources (e.g. talent, knowledge, capital, goodwill) back (below for a schematic representation). Legitimacy, customers, networks, talent pool, supporting services, university Ecosystem Campus Jobs, networks, lowering transaction costs, legitimacy Shared equipment, economies of scale, credibility, social capital, unused knowledge Startup community (e.g. incubator) Economic activity, entrepreneurial culture, networks, investors, technology commercialization Figure 17: schematic representation of start-up incubation in a campus environment A campus is an especially relevant context for start-ups. It can broadly be defined as a geographical area on which research takes place through interaction and collaboration. Several scholars emphasise that campuses hold all the required resources needed to successfully launch start-ups (Link & Scott 2003; Autio et al. 2004; Phan et al. 2005; Salvador & Rolfo 2011; Miller & Acs 2017). Examples of these resources are shared equipment, economies of scale, credibility and relevant networks (Weele et al. 2014). According to the spill over theory of entrepreneurship (Acs et al. 2008; Shu et al. 2014) there is more knowledge developed in regions than can be capitalized on by existing organisations. This knowledge spill over provides unpursued entrepreneurial opportunities for start-ups. Other scholars, however, push back to this enthusiasm and draw attention to possible drawbacks. For example, Oakey Q-campus Background study Background study 36

38 (2007, 2013) argues that campuses also carry risks with regard to IP protection and involuntary spillovers. Furthermore, a review by Siegel et al has not been able to capture the great benefits of campuses for start-ups. They showed negligible returns of being located on a science park. Nevertheless, the general consensus is that the benefits of campuses for start-ups outweigh the drawbacks. Furthermore, start-ups can also provide advantages to the campus they are located on. The start-up community, consisting of start-ups, investors, incubators, etc., can generate economic activity and jobs, which can provide legitimacy in the wider region. Moreover, by absorbing spilled over knowledge they can commercialise unused knowledge. Start-ups furthermore contribute to the formation of networks and attracting relevant stakeholders such as investors and other firms. Finally, start-ups can contribute to an entrepreneurial and energetic culture at the campus. Considering these reciprocal benefits, there is a strong case to make to develop a strong start-up ecosystem at or around a campus. Nowadays, the most popular way of fostering a start-up ecosystem is through business incubation activities. Nevertheless, business incubation is not a simple one-sizefits-all solution. Proper business incubation should be customized to its specific context. Therefore, we first briefly introduce business incubation to then apply this to the specific context of the Q-campus. Start-up incubation Incubation in general entails the supporting of start-ups through providing services and resources. It usually takes place at a physical location at which start-ups interact with each other and other relevant stakeholders. While there are many different practices that an incubator can organise, such as events, workshops, consultancy, progress sessions, creating awareness, lobbying, etc., the main mechanism through which start-ups are supported is through indirect learning in networks (Tötterman & Sten 2005; Eveleens et al. 2016; Hallen et al. 2016). Furthermore, the literature distinguishes between specialised and general incubators. Specialised incubators focus on a specific market or industry. General incubators do not have such a focus. In practice, most incubators have some industry focus, but at the same time keep an open mind on accepting different start-ups. There is a strong case to make that there is an optimum similarity between start-ups in relation to incubation performance (Bollingtoft, McAdam and Marlow). In other words, neither an overly narrow, nor a too broad selection strategy is advisable. Over the last decade, accelerator programmes have become a popular form of business incubation. Riding on the IT wave, accelerator programmes have proven to accelerate the development of startups (Hallen et al., 2017). The main distinguishing feature of these accelerator programmes is the relative short period in which start-ups are supported and the pressure that is created on the start-ups to learn quickly. Apart from this an accelerator provides the same services as other incubation programmes. Q-campus Background study Background study 37

39 Start-up incubation at Q-campus A campus around quantum technology in Delft is exceptional in several ways. In what follows, we characterise the Q-campus context and draw implications for developing a start-up community. First, the technology is still in a relative early stage of development, when considering commercialisation activities. Related to this, the campus is highly dominated by specific academic knowledge. The consequence is that any start-up in this context will suffer from severe legitimacy problems. While the academic field enjoys an appealing scientific image, founding a business on such emergent technology is risky. Potential customers would be hesitant to do businesses with start-ups in this field. Not only do the start-ups have no track record that could boost confidence (by definition), the technology is often not well understood and deemed inappropriate in the context of business of usual. Quantum computing related start-ups would have to deal with this, but would perhaps also benefit from the current increasing attention (or even hype) that is generated for this technology. Furthermore, the world-leading, but narrow knowledge base poses a risk to start-up development. Namely, it hampers the identification of entrepreneurial opportunities, which are typically find on the intersection of fields of knowledge. In particular, to found and develop a start-up, not only technical, but also business and finance knowledge are needed. Therefore, a good start-up strategy would entail forging connections to different bodies of knowledge, including business and finance. Moreover, the timeline for developing quantum technology start-ups is long. Therefore, instead of a 3- month accelerator programme, some patience should be practised before expecting successes. Finally, besides the core quantum technology, there are other aspects of scientific research that can serve as the source of a start-up. Research methods, datasets, research tools, and research networks can be a point of departure for exploring entrepreneurial opportunities. Examples are Delft Circuits that creates components for research tools in quantum technology, and Single Quantum that creates photon detectors. Second, Delft is a region with a long track record in technology-based and science-based venturing. It harbours a rich start-up ecosystem which has spurred many start-ups, some of which have become rather successful. Yes!Delft is a renowned and important player in this ecosystem. The consequence is that before intervening and developing this start-up ecosystem, one needs to carefully assess its current merits. For example, there is already a decreasing number of start-ups that is looking for incubation services, while the number of incubators does not yet significantly decrease. This has led to competition among incubators. It does not mean that there is no room for improvement, especially not considering the specific requirements of quantum technology-related start-ups. Taken together, it makes a lot of sense to use existing facilities of e.g. YesDelft and extend these with specific services that are needed for quantum technology related start-ups. Third, the value chain around quantum computing could be dominated by a few global players that have enormous vested interests in the technology. These major players are specifically focussed on developing the architecture and integration of quantum-bits. As such they will be providing the backbone to the new technological paradigm. Q-campus Background study Background study 38

40 For start-ups, this situation has strategic implications. Particularly, there are some business models that are more likely to be successful than others. Considering the large players in this technologically complex context, we can expect a rich network of specialised suppliers that will cater to the major players. Also, once a dominant architectural design will be established, there will be a host of opportunities for business models that use this standard. Practically, the discussion above warrants attention to a number of aspects, when designing a start-up incubation approach around the Q-campus. Organising the following incubation services, resources, and guidelines would likely improve the chances of successful quantum technology related star-tups. Enhance field building by creating networks and attention around quantum-related start-ups. Set up customised, interdisciplinary networking services Organise and offer long term resources, patience and persistence Maintain an open mind and active sourcing of commercialisable ideas Seek partnerships and complementarities with existing incubation and support activities in the Delft region and beyond Ensure dedicated quantum industry consulting to educate start-ups about strategic considerations in the quantum value chain and industry. Sources Acs ZJ, Braunerhjelm P, Audretsch DB, Carlsson B The knowledge spillover theory of entrepreneurship. Small Bus Econ [Internet]. [cited 2014 Jan 27]; 32: Available from: Acs ZJ, Stam E, Audretsch DB, O Connor A The lineages of the entrepreneurial ecosystem approach. Small Bus Econ. 49:1 10. Aernoudt R Incubators: Tool for Entrepreneurship? Small Bus Econ [Internet]. 23: Available from: Autio E, Hameri AP, Vuola O A framework of industrial knowledge spillovers in big-science centers. Res Policy. 33: Autio E, Kenney M, Mustar P, Siegel D, Wright M Entrepreneurial innovation: The importance of context. Res Policy [Internet]. [cited 2014 May 26]; 43: Available from: Eveleens CP, Van Rijnsoever FJ, Niesten E How network-based incubation helps start-up performance: a systematic review against the background of management theories. [place unknown]: Springer US. Hallen BL, Bingham CB, Cohen SL DO ACCELERATORS ACCELERATE? THE ROLE OF INDIRECT LEARNING IN NEW VENTURE DEVELOPMENT. Seatle, WA. Link AN, Scott JT U.S. science parks: the diffusion of an innovation and its effects on the academic missions of universities. Int J Ind Organ [Internet]. 21: Available from: Miller DJ, Acs ZJ The campus as entrepreneurial ecosystem: the University of Chicago. Small Bus Econ. 49: Phan PH, Siegel DS, Wright M Science parks and incubators: observations, synthesis and future research. J Bus Ventur [Internet]. [cited 2013 Oct 17]; 20: Available from: Q-campus Background study Background study 39

41 Salvador E, Rolfo S Are incubators and science parks effective for research spin-offs? Evidence from Italy. Sci Public Policy. 38: Shu C, Liu C, Gao S, Shanley M The Knowledge Spillover Theory of Entrepreneurship in Alliances. Entrep Theory Pract. 38. Tötterman H, Sten J Start-ups: Business Incubation and Social Capital. Int Small Bus J [Internet]. [cited 2014 Mar 1]; 23: Available from: Weele MA van, Steinz HJ, Rijnsoever FJ van Start-ups down under: How start-up communities facilitate Australian entrepreneurship. In: DRUID Soc Conf 2014 [Internet]. Copenhagen; [cited 2014 Aug 25]. Available from: Q-campus Background study Background study 40

42 5. Campus environment case studies By Elmar Cloosterman Given below is a comparison of four different campuses in The Netherlands that are considered to successful. Information is based on information provided by the organizations managing these campuses, partner organisations, and previously conducted (academic) research. Given the vast discrepancies of the information available between campuses it is hard to provide direct comparison on subjects like financing and governance. Further research in the form of interviews could provide more insight in these subjects WaterCampus Leeuwarden About WaterCampus Leeuwarden is considered the meeting point of the Dutch water technology sector and has the ambition to play a sector uniting role for the rest of Europe as well. WaterCampus has the goal to stimulate cooperation between (inter)national businesses, knowledge institutes and government within the water technology sector. WaterCampus has three managing parters: Wetsus, Centre of Expertise Water Technology (CEW), and Water Alliance. With Wetsus conducting the scientific research at the campus, CEW being the knowledge and innovation centre for applied research and product development and Water Alliance providing the partnership between public and private companies and government institutes. These three organisations thus provide the full ecosystem of the campus in which initial ideas can be researched (Wetsus), tested (CEW) and marketed (Water Alliance). Hence, WaterCampus Leeuwarden is the cooperative organization overviewing these three organisations, with the partner organisations being the owners of the brand WaterCampus. Each of the three managing partner organisations have long lists of partner organizations ranging from knowledge institutes (Universities, Universities of Applied Science, research institutes) and public organisations (Provinces, municipalities, ministries and cooperative organisations such as the SNN) to private organisations (Unilever, Shell, FrieslandCampina, etc.). The managing partners create the campus ecosystem by providing space and facilities to start-ups, scale-ups, incubators and spin-off companies. Q-campus Background study Background study 41

43 Figure 18: WaterCampus its 'ecosystem'. Science = Wetus, applied research = CEW, business = Water Alliance Research and product development activities Research at WaterCampus is divided between two organizations (Wetsus, CEW). With Wetsus focusing on purely scientific research. Currently, the most important research topics for Wetsus range from desalination, biofouling and concentrates to Blue- and CO2-energy. CEW s research agenda is more focused on innovation and practical solutions, with the one of the main topics being the sustainable management of water in The Netherlands. From 2015 to 2017 Wetsus employs a total of 179, 188 and PhD candidates and post-docs 45, respectively, who work collectively on a research program consisting of 22 research themes. In principle all PhD-researchers executing the research projects are employed by the involved know-how institutes and full time seconded in the multidisciplinary Wetsus laboratory in Leeuwarden. CEW, Water Alliance and the overviewing WaterCampus organisation provide no information about fte s and/or product development roadmaps. 44 Number of expected employees based on predictions from Exact number of fte is not provided Q-campus Background study Background study 42