Quantum Software Outlook 2021

IBM continues to dominate the quantum cloud. However access to new more powerful processors will increase competition in the early adopter market. Ideas for user-engagement and education continue to explode with innovation. In the long term, supporting developers is a challenge the conventional software industry understands, but don’t forget that underneath the hood the quantum stack is different. There is room for a radical shake-up by whoever gets this right.

Software proved a vital commercial playing field in the digital revolution. Many expect the same to prove true in the new quantum revolution. A diversity of players are perusing many distinct strategies. Initial communities and ecosystems have genuinely formed.

For a provider the first challenge is in which part of the stack to play and where to partner? Where can core expertise win and where will it always be trumped by scale? Where can rents be defended in the future quantum software value chain?

For early adopters the first challenge is how to allow internal teams to engage and learn? How to turn this initial engagement into organisational knowledge and a realistic roadmap? How to avoid spending time and money locked in on the wrong path?

Quantum Platform-as-a-Service starts to heat up

IBM Quantum

IBM has defined this initial phase of the market, achieving unparalleled penetration with over 265,000 registered users and over 1 billion hardware quantum circuits run on a typical day across 28 quantum processors. Increasingly premium usage dominates and is trending up. Whatever happens next, IBM Q, recently renamed to IBM Quantum, has been a remarkable success.

IBM Quantum Experience launched in 2016 and initially focussed on building awareness by offering a simple graphical web interface where users could create (compose) simple quantum programs (circuits) and then run them on early quantum hardware. IBM built on this success to build-out the world’s first fully-fledged quantum cloud platform for scientific and early industry adopter use. Today the IBM Quantum network partners include commercial majors Daimler, Exxonmobil, JP Morgan Chase & Co, Samsung, Goldman Sachs, Accenture and Boeing, with 130+ members overall. This sits neatly alongside IBM’s overall business services offer.

Qiskit is IBM’s quantum software development kit and contains elements to support all key stack elements: quantum programming at the circuit and pulse level (Terra), common algorithms (Aqua), error characterisation (Ignis) and offline simulation (Aer). It is much the most starred and forked quantum SDK on GitHub.

While others are just starting to build their community, IBM can point to a long sequence of scientific publications already based on work carried out with their platform. They can point to a growing ecosystem of vendors offering libraries and tools compatible with Qiskit. Neither is it just IBM hardware. AQT’s backend integration with Qiskit is a proof of principle demonstration that the framework can support trapped ion processors. IBM emphasises that it is happy for any hardware provider to integrate their backend.

The chasing pack

D-Wave launched Leap, its own quantum cloud service in 2018. This continues to provide a platform of choice for customers focussed on D-Wave’s quantum annealing hardware.

D-Wave Leap provides an environment tailored to implementing business oriented quantum annealing applications. The Ocean SDK provides the development environment. New tools such as the hybrid solver continue D-Wave’s characteristic focus of making it easy for users to configure applications based on their own business problems.

Other hardware players such Rigetti have also offered quantum cloud access to their hardware, often seeking to allow users to access unique features available on their hardware. More recently, platforms from tech majors with deep pockets have gathered steam. The launch of Amazon Braket by AWS and the continued development in preview of Quantum Azure from Microsoft both offer hardware agnostic PaaS visions from the bottom-up. Both leverage their power to offer flexible provisioning of associated conventional cloud computing resources.

Amazon Braket completed its general launch in 2020 offering access to quantum backends from D-Wave, Rigetti and IonQ. The service features its own new SDK and emphasises the benefits of integration with market leading AWS cloud features for deployment and data management. Braket offers a simple Python API for defining the quantum portion of the program but leaves users to build and manage the classical piece. Early blue-chip collaborators have included Volkswagen Fidelity, Amgen, and Enel. Startup and academic partners include Rahko, Qu&Co and IQC.

Azure Quantum from Microsoft has entered a limited preview phase in 2020. It promises access to quantum backends from Honeywell, IonQ, and QCI, plus the Toshiba SBM (for simulated annealing). Users benefit from Microsoft’s established Q# and QDK development tools, as well as the conventional Azure cloud features for deployment and data management. A key feature is the ability of Q# to capture both the classical and quantum program in a single piece of code. This is a natural extension of Microsoft’s traditional strength in developer productivity applications. Partners include 1QBit who provide support for finding solutions across quantum/classical hardware.

Microsoft is able to build on its long involvement with quantum research via its Station Q network. AWS are seeking to catch up ground here. Attracting industry pioneer John Preskill to work (one day a week) at the new AWS Center for Quantum Computing is an eye-catching move. Both players strongly emphasise that they are in the quantum game for the long term.

Google has kept access to their premium hardware closely held, but has now launched an early access program to allow select external parties to access its quantum cloud service. Initial access has mostly been offered to US academic and national lab experts. Quantum software startup PhaseCraft is notably for being the only commercial and only non-US entity invited in as part of the initial stage

Google’s Quantum Computing Service is set to be based initially on access to Sycamore processors in 20-50Q configurations. It leverages the Cirq programming framework and benefits from the OpenFermion and TensorFlow Quantum libraries that have been developed in the course of Google’s own applications work. Code examples from Google’s own experiments are available (ReCirq) and simulator tools (Qsim). Support for certifiable random number generation is a notable feature. All of the tools and libraries are open source, connecting to the Google proprietary Quantum Engine cloud service.

Google emphasises a strategy of limited rather than broad access. This focusses on users that bring key quantum or industry domain expertise to push the limits of what can be achieved with early hardware.

Overall the battle lines in the platform market increasingly reflect the wider positioning of the parent tech majors. The table stakes increasing look to require deep pockets.

Education, education, education

Organisations won’t decide what to do with quantum computers, the people that work in them will. The first step to building a user base is user engagement and education.  

One highlight of 2020 was the inaugural IBM Quantum Computing Challenge events. These combined cloud access to IBM’s quantum hardware with a graded series of coding tasks designed to engage and develop quantum coders with a wide variety of existing skills.

IBM Quantum Challenge – In the first event alone, 1745 people took part, with 574 completing all four exercises correctly. Very much an educational and community building activity the success of early events has led to their repetition. Fact Based Insight believes they are an excellent way for companies to encourage their most talented staff to start their engagement with the quantum journey.

Google may have identified an interesting way to fight back. Quantum Chess from Quantum Realm Games promises to be an intriguing feature of Google’s quantum cloud plans.

Quantum Chess – A novel twist on the classic game – pieces are allowed to make ‘split’ moves placing themselves in superposition. Subsequent moves can create entanglement between different outcomes. When different pieces try to occupy the same square a quantum measurement occurs fixing outcomes.

Q2B 2020 hosted the inaugural Quantum Chess developers’ tournament. After a multi-round event that really captured the interest of the community, Aleksander Kubica (AWS) was crowned the first Quantum Chess world champion.

This is a fun twist on a classic game. It also has the potential to provide an engaging entry point into representing basic quantum circuits in Cirq. The quantum sector needs ways to give newcomers some intuition about how quantum outcomes differ from classical ones. Quantum Chess points to ways this can be done without having to teach everyone a course in quantum physics. This could even have an educational reach down to high school level. Not to be out done, Chinese scientists have developed Quantum Go along similar lines .

Another interesting learning product at the other end of the spectrum is BLACK OPAL from Q-CTRL. This provides strong graphical visualisations of basic qubit operations and circuits. On top of these, users can visually explore the effect of noise and how control techniques help compensate. Again the idea is to give newcomers (and old hands) an intuition for what is going on.

In Europe, QuTech’s Quantum Inspire quantum cloud platform offers access to execution on a variety of simulators and 5Q and 2Q quantum processors (the latter notably the first silicon spin backend on the cloud). Again the emphasis is on user engagement, with easy anonymous basic registration, progressing to supercomputer hosted simulation of up to 31Q.

With more academic roots, ProjectQ is also still actively maintained and offers its own offline simulator and compatibility with IBM Quantum and other backends.

High-end simulation

For serious quantum development, high performance simulation is a key and non-trivial requirement.

Simulators and simulation – Unfortunately the quantum community uses the same word for three different things: conventional computers simulating quantum hardware; quantum hardware designed for special purpose (analogue) simulation of physical systems; quantum software providing (digital) simulation of problems in quantum chemistry and materials science. Here we mean the former!

Even the most powerful conventional systems rapidly start to struggle as the number of qubits to be simulated grows (that is the point after all). However, even when more powerful quantum devices become available, conventional simulation of program components is expected to continue to play a key role in code debugging and validation.

At a superficial level, this may just sound like a conventional high performance compute application. But the complexity of this market is already growing as players optimise for different tasks.

IBM Quantum supports a range of offline and online simulators, currently advertised up to 32Q. Google’s Qsim can simulate around 30Q on a laptop or reportedly up to about 40Q in the Google Cloud. Entrants like Amazon Braket and Azure Quantum place significant emphasis on their ability to flexibly provision conventional cloud hardware to meet user needs. Braket in particular features TN1, a tensor network simulator reported able to simulate circuits with the right structure up to 50Q.

Atos have exploited a niche in the market, offering their Quantum Learning Machine (now QLM E), to those who want a high-power simulator environment without having to rely on cloud access or being tied to a single quantum hardware vendor. Atos already boast of 13+ installations at leading institutes and companies worldwide. QLM E offers simulation to 41Q, with optimisation for variational algorithms up to 30Q.

QuEST is a notable open-source simulator developed at Oxford University. It has been used to simulate up to 38Q on a 49,000 CPU cluster (up to 45Q is believed to be possible if you can provide the right supercomputer).

Chinese cloud offerings currently emphasise the power of their simulator support. Huawei’s HiQ 2.0 (Asia-only for regulatory reasons) simulates up to 42Q . Ablibaba’s AC-QDP claims usability for certain applications even at 50Q+ . OriginQ have recently augmented their own simulator based cloud with access to one of their 6Q quantum processors (with plans to expand to 24Q ongoing) .

Helping the enterprise get to work

As an increasing number of corporate pioneers engage in early quantum computing activities, a number of startups are seeking to provide an enhanced environment for nascent quantum application software development. Enterprise quantum software has arrived.

Some users are already QC experts who want the ability to iterate runs with a minimum of fuss and to port work across platforms and to have access to accelerated simulator tools. Others are data scientists experimenting with quantum for the first time and seeking support with libraries of pre-made algorithms. The players in this sector typically combine a team that brings rare expertise in quantum algorithms, with software tools to help clients get started.

Zapata, QC Ware, 1QBit and Strangeworks all have notable offerings. So far Zapata is by far the best funded, having raised $57m, while QC Ware benefits from the early mover advantage it seized by building-up its popular Q2B conference series.

Orquestra – Zapata’s enterprise platform implements a flexible workflow model to support the multiple runs and iterative execution of the hybrid classical/quantum processing that many quantum solutions are expected to demand. It combines these lessons from conventional enterprise software with a growing library of quantum algorithms. Many organisations will face the challenge of how to turn innovation from expert individuals into organisational know-how and deployment. Orquestra could be well placed to help with this, automating and unifying data handling tasks that would distract experts, and proving a practical repository to operationalise their work. Qrquestra directly integrates with Honeywell, IBM and IonQ, and with Rigetti via AWS Braket.

Forge – QC Ware’s offering focusses on the libraries of customised quantum algorithms it provides for binary optimisation, chemistry simulation, machine learning and Monte Carlo simulation, in addition to the industry targeted software applications it plans to build on these. An interesting feature is the Forge Data Loader designed to accelerate the loading of classical data into a quantum device, a key roadblock for many quantum machine learning applications. Notable clients include Equinor, Airbus, BMW, Goldman Sashs, Aisin and Covestro. Forge integrates with Amazon Braket, and Forge algorithms can now also be run on IBM Quantum. Accelerated simulation is provided by cloud-based Nvidia GPUs.

Quantumcomputing.com – Strangeworks’ community oriented offer also comes from a team with experience writing enterprise code. The style and culture on offer emphasises the desire to help teams share and interact.  In contrast to environments such as Amazon Braket or Azure Quantum, the development environment acts as a hub, offering the ability to manage coding projects directly within all major quantum  frameworks, including Qiskit, Q#, Forest. Cirq, Ocean and Pennylane.

Applications for the future

It is too early for genuine quantum application software to emerge but startups are already positioning themselves for this future market. Balancing quantum algorithms expertise with deep industry insight is the key demand, that latter probably even more vital than in the conventional applications software sector. Having a business model that can stay engaged with the industry and prosper through a much extended development cycle is a key challenge.

Algorithm experts are building libraries of sector specific tools around pilot engagements with flagship clients. Examples include QEMIST by 1QBit, EUMEN by CQC and QAD Cloud by HQS in quantum chemistry; Pennylane by Xanadu and Rahko in quantum machine learning.

Others emphasise synergies with service lines based on conventional AI and data science technologies. Examples include QDL and Multiverse in FS; ProteinQure in drug design. Similarly Qu&Co emphasise their strategic partnership with Schrödinger, a leader in the conventional quantum chemistry software in use today.

Others are leveraging quantum annealing and quantum-inspired techniques to target client benefits now. Examples include POLARISqb and 1QBit’s 1QCloud.

Others are emphasising the strong academic credentials of their algorithms work. This helps secure government funds for cutting edge innovation. Examples include Phasecraft and BEIT.

In the quantum applications market, quantum hype is an ever present danger. Speaking at IQT Europe, Benno Broer (Qu&Co) offered some good advice “It’s a B2B sell. Honesty, integrity and rigor are key to protecting your long term reputation”.

Compilers for the hard core

In contrast to conventional compilers, optimising quantum compilers have to deal with a series of uniquely quantum challenges. Quantum devices have a specific native gate set and restricted qubit connectivity. When mapping an algorithm to a quantum device key choices must be made in qubit placement and routing. There are also unique opportunities. In many cases quantum circuits can be rearranged to perform an equivalent operation with a simpler circuit. Real world quantum devices are noisy, but this noise can be characterised at the individual device level and mitigating action taken at compile time. Technically we’re often actually talking about a transpiling operation, so interoperability is a useful feature.

Two independent quantum compilers stood out from the crowd in 2020, each demonstrating an aspect of the unique know-how that will be necessary to succeed in this market.

t|ket>  – CQC’s flagship compiler scored a notable success when it was credited by Google in its work this year with QAOA on Sycamore . There is a strong mathematical dimension to making a successful quantum optimising compiler.  In  t|ket>’s case this is built on the power of ZX-calculus .  

True-Q – Quantum Benchmark have a strong, respected heritage in the characterisation and mitigation of quantum errors. True-Q features the randomized compiling technique they originally developed to overcome systematic control errors. This has now been expanded with the new error adaptive compiling technique able to compensate for 2Q gate errors . This also holds out the prospect of being able to certify circuit fidelity at compile time.

There are several powerful themes emerging in the compiler market. All build on deep expertise that is in many cases complementary rather than competing.

As the race to implement error correcting codes on early quantum hardware gathers pace, we can expect it to unlock a further wave of compiler innovation.

Rewiring the stack

The emerging shape of the quantum software industry may seem hauntingly familiar, adapting as it does many concepts from the conventional software industry. However, particularly at the lower layers of the stack the differences are pronounced. We are also at a radically earlier phase of hardware development than almost any of us can remember. Is the familiarity of our high level software terminology blinding us to challenges and opportunities lurking below?

Deltaflow.OS from Riverlane is a new full stack quantum operating system that seems set to challenge a number of conventional industry assumptions. A Riverlane led consortium including ARM, Hitachi Europe and five different quantum hardware startups has just secured a £7.6m grant to bring Deltaflow.OS to market.

In contrast to software platforms designed to draw in software activity from early adopters, Deltaflow.OS takes aim squarely at a unified hardware/software R&D market. For this it offers the promise of accelerated development, low latency and the potential for flexible interaction between application and control layers.  It opens back up key lower layers of the software stack typically kept proprietary by current hardware players.

Deltaflow.OS – Quantum processors are typically driven by a conventional host processor. Between the two most envisage a network of global and local control nodes. Deltaflow.OS streamlines the task of getting custom code onto control nodes implemented by FPGAs, emphasising reduced instruction set implementations that are easier to debug. This approach promises to reduce R&D cycle time. It also uses a distributed rather than hierarchical network node concept, including the potential for flexible application and control stack interaction. These features promise to minimise runtime latency.

Deltaflow.OS has now launched its initial module, an integration with ARTIQ (a popular trapped ion control system, for example ).  However it’s important to realise that this initiative is much wider in scope and already benefits from the involvement of the unusually broad mix of qubit technology startups currently active in the UK (two flavours of superconducting qubit, two flavours of ion trap, silicon and photonic architectures are all represented). It also builds on (hard) lessons learnt during phase 1 of the UK NQT programme.

Deltaflow.OS isn’t incompatible with other cloud and enterprise software initiatives, but it does represent a different view of where maximal value will be created. No one knows how the quantum computing sector will evolve. We could see leading technologies continue to pull in front and efforts focus around a small number of leading programmes. However, in the corner of the multiverse where a diversity of different qubit technologies and startups compete to implement error correction, Deltaflow.OS could find a distinct niche. In another corner of the multiverse where commercial NISQ applications emerge but rely on coding gymnastics to squeeze the most out of early hybrid or bespoke solutions, Deltaflow.OS could do very well indeed.

Quantum software endgame

Whatever’s happening underneath with quantum hardware, the developer’s ‘user’ perspective will increasingly include the quantum software stack. While we have many intricate measures for quantum hardware performance, we still have little that addresses what it’s like to use these systems end-to-end on the cloud. Tom Lubinski (Chair of the QED-C Standards Committee) points out “Many users are clamouring for something, anything that helps them decide where they should invest their precious time and money”.

In the long term, many believe that the overall developer experience will have to be wrapped up into a much simpler whole. Speaking at Q2B, Eric Schmidt (former Google CEO) observed  “Eventually what’s going to happen  is that people will be writing in the equivalent of Python and PyTorch again, 8, 9 or 10 years from now”.  How quickly this transition will be able to proceed is a question that will be tested again and again in the coming years.

To watch in 2021

  • Quantum Supremacy on the cloud – Google has not so far managed to repeat its quantum supremacy experiment in routine (automatically calibrated) running of Sycamore. The first cloud offering to enable users to run their own computational quantum supremacy proof will claim significant bragging rights in this market.
  • Cloud volume – Quantum volume is a more rounded measure of early quantum processor capability. If trapped ion offerings can build a lead here, expect a big pull through for the platforms that offer access.
  • Cloud logos – Expect leading platforms to compete on the quality of their blue-chip partner list. Watch for organisations with the intellectual bandwidth to drive substance behind these initiatives.
  • Cloud Benchmarks – Watch as standard ‘real problem’ benchmarks develop. D-Wave’s new Qiskit plugin aims to make it easy to benchmark an important class of optimisation problems across any backend supported by Qiskit as well as its own quantum annealing hardware. Watch out for sparks.
  • Agnostic platforms – Hardware agnostic quantum cloud platforms are on the rise backed by deep pocketed tech majors. How will own-brand full stack platforms adapt?
  • Engagement – IBM Quantum has a daunting lead in user engagement and education which is unlikely to be quickly overturned. Will we see a competitor start to make any progress in clawing this back?
  • Education – Watch out for the launch of products with a specific focus on the education market.  BLACK OPAL 2.0 is expected to cover discovering quantum mechanics to executing algorithms, with content from Chris Ferrie (author of Quantum Physics for Babies). Quantum Chess 2.0 promises to include puzzles calling for quantum solutions. Can we expect to see an AI opponent from Google’s DeepMind?
  • Compilers – Watch out for new practical optimising features. Will compilers aim to transpile nicely together or will they slug it out?
  • Watch the Horizon – Horizon Quantum Computing is opening its proprietary compiler and software developer tools to early users. These have the ambitious goal of allowing developers to automatically construct quantum algorithms based on programs written in conventional languages such as Matlab. Will this platform allow a broader range of software developers, without quantum-specific knowledge, to bring their skills to the table?
  • Watch the Network – Watch out for the progress of Aliro Quantum’s  first two products Q.Compute and Q.Network. Will these catch the buzz growing around  quantum networks?
  • Simulators – High performance simulation will remain key to understanding, testing and validating quantum software. Watch out for optimised simulator performance as a key differentiator at the top end.
  • Quantum Chess vs Quantum Go – Quantum cloud platforms from the Chinese majors and startups are increasingly seeking to replicate the user engagement journey pioneered in the West. There will be similarities, but also cultural differences in terms of what works to build momentum.

References

[1]
QuTech, “Quantum Inspire,” QuTech. [Online]. Available: https://qutech.nl/research-engineering/quantum-computing/quantum-inspire/. [Accessed: 14-Dec-2020]
[1]
“Welcome to True-QTM — True-QTM.” [Online]. Available: https://trueq.quantumbenchmark.com/. [Accessed: 16-Dec-2020]
[1]
“A first look at Deltaflow.OS® at the UK National Quantum Technologies Showcase 2020,” Riverlane, 16-Nov-2020. [Online]. Available: https://www.riverlane.com/news/2020/11/a-first-look-at-deltaflow-os-at-the-uk-national-quantum-technologies-showcase-2020/. [Accessed: 16-Dec-2020]
[1]
“Commercial breakthrough following success of Deltaflow.OS trials,” Riverlane, 01-Sep-2020. [Online]. Available: https://www.riverlane.com/news/2020/09/commercial-breakthrough-following-success-of-deltaflow-os-trials/. [Accessed: 16-Dec-2020]
[1]
“Strangeworks | Quantum Computing Platform,” Strangeworks. [Online]. Available: https://strangeworks.com/. [Accessed: 16-Dec-2020]
[1]
“QC Ware Corporation,” GitHub. [Online]. Available: https://github.com/qcware. [Accessed: 16-Dec-2020]
[1]
“Zapata Computing, Inc.,” GitHub. [Online]. Available: https://github.com/zapatacomputing. [Accessed: 16-Dec-2020]
[1]
“QuEST – Quantum Exact Simulation Toolkit.” [Online]. Available: https://quest.qtechtheory.org/. [Accessed: 16-Dec-2020]
[1]
“Atos takes the most powerful quantum simulator in the world to the next level with Atos QLM E,” Atos, 23-Jun-2020. [Online]. Available: https://atos.net/en/2020/press-release/general-press-releases_2020_06_23/atos-takes-the-most-powerful-quantum-simulator-in-the-world-to-the-next-level-with-atos-qlm-e. [Accessed: 16-Dec-2020]
[1]
“Amazon Braket tensor network simulator supports 50-qubit quantum circuits,” Amazon Web Services, Inc. [Online]. Available: https://aws.amazon.com/about-aws/whats-new/2020/12/amazon-braket-tensor-network-simulator-supports-50-qubit-quantum-circuits/. [Accessed: 16-Dec-2020]
[1]
“ProjectQ-Framework,” GitHub. [Online]. Available: https://github.com/ProjectQ-Framework. [Accessed: 16-Dec-2020]
[1]
“BLACK OPAL®,” Q-CTRL. [Online]. Available: https://q-ctrl.com/products/black-opal/. [Accessed: 16-Dec-2020]
[1]
“IBM Quantum Challenges,” IBM Quantum Experience. [Online]. Available: https://quantum-computing.ibm.com/challenges. [Accessed: 16-Dec-2020]
[1]
“quantumlib,” GitHub. [Online]. Available: https://github.com/quantumlib. [Accessed: 16-Dec-2020]
[1]
microsoft/Quantum. Microsoft, 2020 [Online]. Available: https://github.com/microsoft/Quantum. [Accessed: 16-Dec-2020]
[1]
aws/amazon-braket-sdk-python. Amazon Web Services, 2020 [Online]. Available: https://github.com/aws/amazon-braket-sdk-python. [Accessed: 16-Dec-2020]
[1]
“D-Wave Systems Inc.,” GitHub. [Online]. Available: https://github.com/dwavesystems. [Accessed: 16-Dec-2020]
[1]
“Qiskit,” GitHub. [Online]. Available: https://github.com/Qiskit. [Accessed: 16-Dec-2020]
[1]
“Origin Quantum Cloud Platform.” [Online]. Available: http://www.originqc.com.cn/en/website/classIntroduction.html?bannerId=57. [Accessed: 16-Dec-2020]
[1]
“Huawei HiQ quantum computing software cloud platform.” [Online]. Available: https://hiq.huaweicloud.com/. [Accessed: 16-Dec-2020]
[1]
C. Huang et al., “Alibaba Cloud Quantum Development Platform: Surface Code Simulations with Crosstalk,” arXiv:2002.08918 [quant-ph], Feb. 2020 [Online]. Available: http://arxiv.org/abs/2002.08918. [Accessed: 02-Dec-2020]
[1]
C. Cantwell, “Quantum Chess: Developing a Mathematical Framework and Design Methodology for Creating Quantum Games,” arXiv:1906.05836 [quant-ph], Jul. 2019 [Online]. Available: http://arxiv.org/abs/1906.05836. [Accessed: 20-Nov-2020]
[1]
L.-F. Qiao et al., “Quantum Go Machine,” arXiv:2007.12186 [physics, physics:quant-ph], Jul. 2020 [Online]. Available: http://arxiv.org/abs/2007.12186. [Accessed: 02-Dec-2020]
[1]
F. Arute et al., “Quantum Approximate Optimization of Non-Planar Graph Problems on a Planar Superconducting Processor,” arXiv:2004.04197 [quant-ph], Jul. 2020 [Online]. Available: http://arxiv.org/abs/2004.04197. [Accessed: 13-Nov-2020]
[1]
L. Egan et al., “Fault-Tolerant Operation of a Quantum Error-Correction Code,” arXiv:2009.11482 [quant-ph], Sep. 2020 [Online]. Available: http://arxiv.org/abs/2009.11482. [Accessed: 27-Nov-2020]
[1]
A. Erhard et al., “Characterizing large-scale quantum computers via cycle benchmarking,” Nat Commun, vol. 10, no. 1, p. 5347, 2019, doi: 10.1038/s41467-019-13068-7. [Online]. Available: http://arxiv.org/abs/1902.08543. [Accessed: 22-Apr-2020]
[1]
R. Duncan, A. Kissinger, S. Perdrix, and J. van de Wetering, “Graph-theoretic Simplification of Quantum Circuits with the ZX-calculus,” arXiv:1902.03178 [quant-ph], Mar. 2020 [Online]. Available: http://arxiv.org/abs/1902.03178. [Accessed: 22-Apr-2020]

Quick Navigation

Quantum Outlook 2021Hardware / Algorithms / Software / Internet / Sensing / Landscape

David Shaw

About the Author

David Shaw has worked extensively in consulting, market analysis & advisory businesses across a wide range of sectors including Technology, Healthcare, Energy and Financial Services. He has held a number of senior executive roles in public and private companies. David studied Physics at Balliol College, Oxford and has a PhD in Particle Physics from UCL. He is a member of the Institute of Physics. Follow David on Twitter and LinkedIn

Leave Comment