sign in sign up
<<
Home , Quantum computing, Sycamore processor

Quantum in the UK today Dr Rupesh Srivastava, User Engagement May 2021

For BCS Berkshire For BCS Berkshire Members What we’ll cover

1. What is Computing and why the excitement?

2. The status and outlook for & Simulation

3. The UK QT Programme

4. How to engage with Quantum Computing

Question: When will Quantum Computing make an impact? For BCS Berkshire Members “Quantum is a radical departure in information technology, more fundamentally different from current technology than the digital is from the abacus.” W. D. Phillips Nobel Laureate 1997

Saunpan Abacus Modern Laptop Computer A recent history of science and For BCS Berkshire Members

19th 20th 21st Quantum 2.0

ENIAC

Babbage Difference Engine Colossus at Bletchley Park Summit QCS Hub Research For BCS Berkshire Members Classical influences the design of the latest chips

Some components may use quantum principles in their operation – but the chip does not use quantum for computation 32-core AMD Epyc (2017) 19,200,000,000 transistors (14 nm)

BREAKING NEWS 2021 For BCS Berkshire Members Why build a quantum computer?

The first question is, What kind of computer are we going to use to simulate physics?

But the physical world is quantum mechanical, and therefore the proper problem is the simulation of quantum physics.

Can you do it with a new kind of computer, a quantum computer? Nobel laureate … It’s not a , but a machine of a different kind. For BCS Berkshire Members Why build a quantum computer?

Opportunity New era of discoveries

Motivation Moore’s Law can’t continue forever

Timing Because now we can!* * 40 years later For BCS Berkshire Members Timing – is key

Theory Experiment You need a happy confluence of theory, experiment and engineering Engineering For BCS Berkshire Members Timing – what about now?

Theory Experiment It’s just possible to engineer a quantum For BCS Berkshire Members Quantum Computing and

Use and manipulate phenomena of quantum physics

• Superposition: • Qubits being in multiple states at the same time, e.g. ‘0’ and ‘1’

• Entanglement: • Grouped behaviour of multiple ‘qubits’

• Interference • Utilising wave properties to amplify the “right” answers and (Source: Blackett Review, 2017) cancel out the “wrong” ones For BCS Berkshire Members Quantum Computing and Qubits

In order to simulate N qubits requires 2N classical

N 2N Comment

3 8 easy! 6 20 ~1x10 laptop 15 50 ~1x10 supercomputer

90 (Source: Blackett Review, 2017) 300 ~2x10 > atoms in known universe For BCS Berkshire Members Illustrating superposition with

A ‘’ is a particle of light

1 0 Which path?

Quantum physics says that even a single photon can pass through both slits

A single photon encodes a “” – a superposition of “0” + “1” For BCS Berkshire Members Why the excitement?

• A Quantum Computer’s power doubles with every qubit added • Speedups beyond the capability of future digital computing • Able to solve ‘hard problems’ • Able to mimic real world quantum systems • Expected high impact on most industry sectors, national economies, national security and research and discovery • Estimated impact, $450-800bn operating income/yr* *(Source: Where will Quantum Computer Create Value – And When?, BCG, 2019)

Source: The Business Potential of Quantum Computing, BCG @ Q2B, Dec 2019 For BCS Berkshire Members A variety of application areas… starting with M

• Medicine • • Materials • Mobility • Molecules • Manufacturing • Modelling & Simulation • Mass Communication • Money Markets • Mysteries of the universe For BCS Berkshire Members Application areas (cont’)

(Source: The Business Potential of Quantum Computing, BCG, 2019) For BCS Berkshire Members A variety of ways to build quantum

Superconducting Solid Cold Nano Ion-traps circuits state atoms wires

Oxford Quantum Oxford Ionics Duality Photonics Quantum Cold Quanta Microsoft Circuits Universal Quantum ORCA Computing Intel Atom Computing IBM IonQ PsiQuantum Quantum Brilliance Pasqal Atos Xanadu Rigetti Honeywell QuiX Intel Alibaba MATERIALS IN SUPERCONDUCTING QUANTUM BITS

box (a ) by Nakamura and co-workers 8 9 MATERIALS IN SUPERCONDUCTINGin 1999. InQUANTUM 2002, Vion BITS et al. developed the Trapped-ion qubitsTrapped for -quantumion qubits for computing quantum computingquantronium qubit (a modifi ed charge qubit) with a T2 time of hundreds of nanoseconds David Lucas, NQIT OxfordFor BCS Berkshire Members David Lucas, NQIT Oxford based on the concept of design and operation Quantum computing: some qubit technologies at fi rst-order noise-insensitive bias points. Example ion traps: box (a charge qubit)Burkard by Nakamura et al. and 1 0 co-workerselucidated the importance Example ion traps: 1mm in 1999. 8 In 2002,of Vion symmetry et al. 9 indeveloped qubit designs, the which in con- Superconducting Superconducting1mmIons circuits Diamond quantronium qubit (ajunction modifi ed with charge Bertet qubit) et al. with brought persistent- current fl ux qubit coherence times into the a T2 coherence time of hundreds of nanoseconds 1 1 IBM based on the conceptfew of microseconds design and operation range. In 2005–2006, 12 13 at fi rst-order noise-insensitiveIthier et al. and bias Yoshihara points. et al. extensively measured the noise properties of quantro- Burkard et al. 1 0 elucidated the importance 5mm Microfabricated surface traps with nium and fl ux qubits, respectively, to better integrated microwaveof circuitry symmetry in qubit designs, which in con- Superconducting circuitsMacroscopicMacroscopic understand the sources of decoherence. The Oxford junction with Bertet et al. brought persistent- 5mm “blade”Microfabricated trap surface traps with ~1 cm“” qubit developed by Schoelkopf integrated microwave circuitry current fl ux qubitand coherence co-workers times signifi into cantlythe reduced the IBM Macroscopic ion trap Key features: few microseconds range. 1 1 In 2005–2006, • Electric circuits made from charge sensitivity+ of the Cooper pair box • room temperature operations Ithier et al. 12 and Yoshiharasingle Ca etion al. 13 extensively superconducting materials• high -vacuum 9-qubit “memory register” by adding a shunt interdigitated capacitor, • 10μm – 100μm feature sizes measured the noisewhich properties would later of quantro-bring microsecond times • Fabricated using semiconductor Key features: • optical () and electronic control nium and fl ux qubits,to the respectively, cavity-QED (quantumto better electrodynamic) industry cleanroom techniques + • room temperature operations understandsingle the Ca sourcesionarchitecture. of decoherence. 14 , 15 An MIT/NEC The group increased Oxford • high-vacuum 9-qubit “memory register” “transmon” qubit Tdeveloped above 20 µby s withSchoelkopf a persistent-current fl ux • Operated at very Microfabricatedtemperature 9 qubit 2 • 10μm – 100μm feature sizes and co-workers qubit signifi incantly a 2D reduced geometry the using dynamical to make them quantum “chip” trap “memory register” • Electric circuits made• optical from (laser) and electronic control charge sensitivitydecoupling of the Cooper sequences, pair 1 6 box and the 3D-cavity superconducting •materialsControlled using microwaves by adding a shuntapproach interdigitated developed capacitor, at Yale 1 7 and now used Oliver & Welander ~ 2 nm Figure 1. Superconducting qubits and their coherence. (a) The three fundamentalwhich would laterby bring several microsecond groups has timesfurther increased this • Fabricated using •semiconductorIn ~20 years of development, MRS Bulletin 2013 coherencesuperconducting has improved qubit modalities: >1000x charge, f ux, and phase. Each includes oneto theor more cavity-QED time (quantum to around electrodynamic) 100 µ s. 1 8 More recently, improve- industry cleanroom techniquesJosephson junctions (shown in red). Illustration by Corey Reed, adapted from Reference 3. architecture. 14 , 15 Anments MIT/NEC in substrate group increasedpreparation and materials (b) The Josephson junction acts as both an inductor, L J , and capacitor, C J . External inductors Superconducting Circuits for Quantum Computing — Peter Leek choices, addressing issues elucidated by the 3D • Operated at very low temperatureand capacitors, L ext and C ext , can be added to modify the qubit’s potentialT 2energy above 20 µ s with a persistent-current fl ux to make them quantumlandscape and reduce sensitivity to noise. (c) Because the Josephson junctionqubit inductance in a 2D geometrygeometries, using have dynamicalled to improved coherence is nonlinear, the qubit potential is anharmonic. The qubit comprises the two-lowest states 1 9 , 2 0 of 2D 1 6 geometries. T h i s fi ve-orders-of- and is addressed at a unique frequency, f 01 . (d) 15 years of progress in qubit coherdecouplingence times, sequences, and the 3D-cavity • Controlled using microwavesreminiscent of Moore’s Law for semiconductor electronics. On average, the doublingapproach rate for developed magnitude at Yale improvement 1 7 and now used in the primary single- coherence times in superconducting qubitsOliver is about & Welanderonce per year. Improvements have been qubit metric can be justly termed a “Moore’s • In ~20 Figure years 1. Superconducting of development,driven qubits by both and new their device coherence. designs (a)and The materials three fundamentaladvances. Qubit modalitiesby represented several groups has further increased this MRS Bulletin 2013 Law for quantum coherence,” 2 1 approaching coherencesuperconducting has improved qubit modalities:include >1000x charge, charge, 8 quantronium, f ux, and phase. 9 f ux, 83 Each 2D transmon, includes 15 one f uxonium, or more 84 and 3D transmontime to 17 qubits.around 100 µ s. 1 8 More recently, improve- Josephson junctions (shown in red). Illustration by Corey Reed, adapted from Reference 3. levels required for a certain class of fault-tolerant ments in substrate preparation and materials 2 2 – 2 4 (b) The Josephson junction acts as both an inductor, L J , and capacitor, C J . External inductors codes. I n a d d i t i o n , and capacitors,Superconducting L and C , Circuitscan be for added Quantum to modifyComputing the — qubit’s Peter Leek potential energy choices, addressing issues elucidated by the 3D ext ext generally much larger than the dilution refrigerator tempera- the control of single 2 5 – 2 9 a n d c o u p l e d 3 0 , 3 1 qubits has also advanced, landscape and reduce sensitivity to noise. (c) Because the Josephson junction inductance geometries, have led to improved coherence ture, T ( i . e . , hf > > k T , w h e r e h is Plank’s constant and k with reports of gate fi delities as high as 99.85%. 28 , 29 is nonlinear, the qubit potential is anharmonic.01 The qubitB comprises the two-lowest states B of 2D geometries. 1 9 , 2 0 T h i s fi ve-orders-of- and is addressed at a unique frequency,is the Boltzmann f 01 . (d) 15 years constant). of progr Hence,ess in qubit the coherdominantence times, T1 -process There is a general consensus within the community that reminiscent of Moore’s Law forhere semiconductor is energy electronics. emission toOn the average, environment the doubling coinciding rate for with magnitudeunderstanding improvement and further in the primarymitigating single- sources of decoher- coherence times in superconducting qubits is about once per year. Improvements have been the qubit relaxing from the to the . qubitence metric in superconducting can be justly termedqubits ( Figure a “Moore’s 3 ) i s o n e k e y t o m o r e - driven by both new device designs and materials advances. Qubit modalities represented 2 1 include charge, 8 quantronium, 9The f ux, second 83 2D transmon, is the 15transverse f uxonium, 84 relaxation and 3D transmon rate (decoherence 17 qubits. Lawadvanced for quantum circuits coherence,” and systems engineering.approaching Indeed, coherence levels required for a certain class of fault-tolerant rate) Γ2 = 1 / T2 , which characterizes the time T2 over which the times should be made as long as possible, as exceeding 2 2 – 2 4 device remains phase coherent; it is related to both the pure quantumthe thresholds error correction for quantum codes. error I n a d correction d i t i o n , will consid- generally much larger than the dilution refrigerator tempera- the control of single 2 5 – 2 9 a n d c o u p l e d 3 0 , 3 1 qubits has also advanced, dephasing time T φ (i.e., a blurring of the relative phase in a erably reduce the resource requirements. Both T1 and T φ 28 , 29 ture, T ( i . e . , hf01 > > k quantum B T , w h e r esuperposition h is Plank’s state) constant and Tand 1 (i.e., k B losingwith the reports excited- of gateare fi delities related as to high the environmentalas 99.85%. noise seen by the qubit, as is the Boltzmann constant).state component Hence, the of dominant a quantum T superposition1 -process state There is a phase- is a generalcharacterized consensus by within a spectral the density, community S ( f ), that and much is known here is energy emissionbreaking to the process).environment coinciding with understanding andabout further this mitigating noise. For sourcesexample, of inhomogeneous decoher- dephasing the qubit relaxing from the Spectacular excited stateimprovement to the ground in the state.capabilities ence of insupercon- superconductingarises qubits from broadband,(Figure 3 ) low-frequency i s o n e k e y t o m(e.g., o r e - 1/f -type) noise in The second is the transverseducting qubits relaxation over the rate past (decoherence decade has brought advanced these qubits circuits andthe systemscharge, flengineering. ux, and critical Indeed, current. coherence However, although it is

rate) Γ2 = 1 / T2 , which characterizesfrom a scientifi the c time curiosity T2 over to thewhich threshold the oftimes technological should be madeconsistent as long with asa bath possible, of two-level as exceeding fl uctuators (or clusters of device remains phasereality. coherent; 7 (see it Figureis related 1d toand both Figure the pure 2 ) Many the individual thresholds ef- forfl uctuators), quantum errorits microscopic correction origin will is consid- not yet well understood. forts contributed to this improvement, starting with the dem- Energy relaxation occurs due to noise at the qubit frequency, dephasing time T φ (i.e., a blurring of the relative phase in a erably reduce the resource requirements. Both T1 and T φ onstration of nanosecond-scale coherence in a Cooper pair S (f ) , a n d d e s i g n m o d i fi cations can change the sensitivity state) and T 1 (i.e., losing the excited- are related to the environmental01 noise seen by the qubit, as state component of a quantum superposition state is a phase- characterized by a spectral density, S ( f ), and much is known

breaking process).,D:0335 7DDD 20183 2 /8C85.53about0 this noise. 1927018350C08:01:0For example, MRS BULLETIN inhomogeneous • VOLUME 38 • OCTOBER dephasing 2013 •7DDD www.mrs.org/bulletin 20183 2817 Spectacular 738 improvement in the capabilities of supercon- arises from broadband, low-frequency (e.g., 1/f -type) noise in ducting qubits over the past decade has brought these qubits the charge, fl ux, and critical current. However, although it is from a scientifi c curiosity to the threshold of technological consistent with a bath of two-level fl uctuators (or clusters of reality. 7 (see Figure 1d and Figure 2 ) Many individual ef- fl uctuators), its microscopic origin is not yet well understood. forts contributed to this improvement, starting with the dem- Energy relaxation occurs due to noise at the qubit frequency,

onstration of nanosecond-scale coherence in a Cooper pair S (f 01 ) , a n d d e s i g n m o d i fi cations can change the sensitivity

,D:03357DDD 20183 2 /8C85.5301927018350C08:01:0 MRS BULLETIN • VOLUME 38 • OCTOBER 2013 •7DDD www.mrs.org/bulletin 20183 2817 738 For BCS Berkshire Members The available NISQ’s • IBM Q, 21 QC’s online (free or paid), up to 65 qubits – roadmap to 1000+ by 2023 • Google Sycamore, access on request, up to 54 qubits • Rigetti, access on request or via AWS Braket, up to 32 qubits • IonQ, access on request or via AWS Braket, up to 11 qubits • Honeywell, access TBA, up to 10 qubits • QuTech’s Quantum Inspire, open access with up to 5 qubits 30+ new ventures worldwide building quantum computer systems … including 6 in the UK Announced: • UK National Quantum Computing Centre • Finland VTT to acquire a National Quantum Computer • Germany (€2b, France (€1.8b), India ($1b) For BCS Berkshire Members IBM Q as an example

(Source: IBM, 2020) What about usefulness: Google’s

• Google used its 54* qubit Sycamore • Non-error-corrected quantum computer (i.e. a NISQ) • In collaboration with NASA ran a well defined computational task (RNG) • The experiment took 200 seconds • Google determined that it would take the world’s fastest supercomputer 10,000 years to produce a similar output * with 53 qubits working • IBM’s response included a proposal that would use Summit, the 2nd most powerful supercomputer, using 250 petabytes of hard-disk space running for 2.5 days

For BCS Berkshire Members For BCS Berkshire Members But not so fast …

• QC’s and Qubits are fragile, difficult to control, difficult to read, full of errors, slow and expensive

• Quantum Error Correction (QEC) enables a Universal Fault Tolerant Quantum Computer (UFT) with ‘logical qubits’ but requires large numbers of physical qubits • The QEC ‘overhead’ depends on the quality of the physical qubits

(Source: IBM, 2018) For BCS Berkshire Members The Development Tools and Environments

• IBM Quantum Experience and • and ProjectQ, promoted by Google AI • Rigetti’s Forest platform with Quil • Amazon’s Braket multi-vendor platform and environment • Microsoft’s Quantum Development Kit including Q# • Atos/STFC Quantum Learning as a Service • QuEST open source Quantum Simulation Kit from Oxford • Deltaflow, River Lane’s cross platform • Quantum Inspire platforms from QuTech in Delft • Zapata’s Orquestra, a unified quantum operating environment Source: qiskit.org • Xanadu’s Pennylane and Strawberry Fields python libraries • Baidu’s Paddle Quantum and many more … For BCS Berkshire Members One more thing …

Relies on factoring prime numbers.

What two prime numbers produce • 15 • 82,249 • 282,589,933 (24,862,048 digits)

Source: Security Magazine, September 22, 2020 For BCS Berkshire Members One more thing … encryption

• NISQ’s are not a threat to most current encryption methods – but UFT Quantum Computers are! • Key risks: • Current data can be stored for future decryption • Re-encrypting existing data takes time • Post quantum – is it truly resilient? • From the National Cyber Security Centre on Quantum Safe Cryptography (QSC):

Early adoption of non-standardised QSC is not recommended. The NCSC recognises the serious threat that quantum computers pose to long-term cryptographic security. QSC using standards-compliant products is the recommended mitigation for the quantum threat, once such products become available. https://www.ncsc.gov.uk/whitepaper/quantum-safe-cryptography For BCS Berkshire Members

Quantum Technologies R&D Source: CIFAR – A QUANTUM REVOLUTION For BCS Berkshire Members Quantum Technologies in the UK

• Long history of outstanding science and QT pioneers • A world-class advanced technology sector • A UK National QT Programme since 2013 • Blackett Report in 2017 • HoC Science & Technology Committee Report, 2018 • UK QTP Phase 2 started in 2019 For BCS Berkshire Members UK National QT Programme, 10 years, £1 billion

• Four Hubs (Sensing, Imaging, Communications, Computing & Simulation) • Skills and Training programme • Industrial Strategy Challenge Fund for Industry-led Innovation • National Quantum Computing Centre • Strategic Partnerships and International participation • 10-year Quantum Strategy led by UK Government For BCS Berkshire Members 2014-2019 Achievements

• Photonically-networked ion trap architecture: node-node connectivity demonstrated with a world-leading combination of rate and fidelity • New benchmarks for speed and precision of operations • Modular quantum optical circuits for processing and simulation • Unique deterministic NV centre-writing capability • Unique superconducting qubit architecture • Blind quantum computing and verification concepts • World’s fastest emulator, QuEST • Verifiable Quantum Random Number Generator • Responsible Research and Innovation studies in QC • Vibrant network of over 100 companies engaged • Encouraged and supported 7 spinouts • International advocacy and industrial engagement • Skills and training in quantum science, technology and entrepreneurship For BCS Berkshire Members 2019 – 2024 The QCS Hub

• The UK National Quantum Computing & Simulation Hub • Five-year research and technology programme with £25m funding, following on from the 2014-2019 NQIT Hub • Objective: to create a economy in the UK • Focus on QC&S technologies, both near-term (NISQC) and long-term (UFTQC) • 17 participating universities, 43 Co-Investigators, led by Oxford • 28 companies and organisations offering support • Supporting the UK National Centre for Quantum Computing For BCS Berkshire Members QCS Hub Partners Academic partners Industrial partners

Airbus Defence & Space National Cyber Security Centre BP National Physical Laboratory BT Oxford Instruments Ltd Cambridge Quantum Computing Limited Oxford Quantum Circuits Creotech Instruments SA Oxford Sciences Innovation D Wave Systems QinetiQ Defence Science & Tech Lab Quantum Machines Element Six Quantum Fraunhofer Institute QxBranch GlaxoSmithKline Rigetti & Co Gooch & Housego Rolls-Royce Heilbronn Institute for Mathematical Research The Alan Turing Institute IQE Ltd Trakm8 Ltd Johnson Matthey M Squared For BCS Berkshire Members QCS Hub Work Programme For BCS Berkshire Members

QCS Hub Context Community Public • The Hub creates the building blocks for the NQCC Policy-makers and industry to scale and exploit Expert advice Quantum Computing Outreach • Industry Engagement with all parts of the value National Quantum & Simulation Hub Computing Centre chain: enabling technologies, system integrators, Skills Devices Skills Research (NQCC) application developers and users Component Training Knowledge/IP improvement Engagement Scaling and • Partnerships with IBM, Google a.o. Architectures building QIP Systems Technologies • Encouraging quantum readiness (QRP) and Software IP/Knowledge Demonstrators Skills/Education quantum literacy in industry and with the public Technologies Applications/Use cases • Encourage quantum literacy through outreach (Quantum City) • Support all parts of the UK National Quantum Technologies Programme: Technology Hubs, Skills Industry and Commerce Hubs, Innovate UK ISCF, KTN Quantum SiG, NPL, The National Quantum Computing Centre, policy makers and sponsors including Dstl and NCSC For BCS Berkshire Members How the QCS Hub can help We collaborate with suppliers, integrators and developers, prospective users, entrepreneurs and investors, and regulators: • Partnership Resource Funding for collaborative projects • Publish Reports and run a Quantum Readiness Programme • Partnerships with system vendors, e.g. IBM Quantum • Access to the research community • Public outreach through Quantum City, a UKNQTP collaboration

To get involved, please contact the engagement team at qcshub.org For BCS Berkshire Members User Engagement – Mission

Develop and nurture the quantum computing ecosystem for UK prosperity For BCS Berkshire Members User Engagement – Ecosystem*

Skills: • Good with people + technical + analytical • Communication – verbal + written • Big picture – but dive in to detail • Curious • Flexible • Entrepreneurial • Innovative • Diplomacy

*Regional, National, International For BCS Berkshire Members The UK Quantum Industry Landscape* (For illustration only, no relationships implied)

Hardware Software Security

Systems

Users

* Not a detailed list For BCS Berkshire Members Publicly announced developments Hardware

• £10m project led by Rigetti UK to build quantum computer based on superconducting circuits

• Consortium includes: Oxford Oxford Instruments: ProteoxLX dilution refrigerator Instruments, University of Edinburgh, (Base temperature as low as 7mK) Phasecraft, and Standard Chartered Bank.

Rigetti quantum processor For BCS Berkshire Members Publicly announced developments Software

• £7.6m project led by Riverlane to build DeltaFlow.OS – a quantum computing operating system

• Consortium includes: SeeQC, Hitachi Europe, Universal Quantum, Duality Quantum Photonics, Oxford Ionics, Oxford Quantum Circuits, Arm and the National Physical Laboratory. For BCS Berkshire Members Publicly announced developments Commercial

IonQ IPO US $2bn! For BCS Berkshire Members The Technological Landscape

We’re not starting from the same roots as classical computing!!

• Quantum computers – emerging • Cloud – democratise access to quantum computers • Powerful classical computers • Machine learning • Neuromorphic computing

Threats or opportunities? “How do you bridgeFor BCS the Berkshire Quantum Divide?”Members The Education Landscape

How old do you need to be to learn quantum computing?

• Qubit by Qubit

• 10,000 students across 125 countries • Middle School, High School and above!

• Q-Munity

• Started by a 14 year old

• Taught herself quantum computing and created an to detect Parkinson’s using speech data

• Many online courses and tutorials – but can you be hired? For BCS Berkshire Members How to engage with quantum computing

• Talk to us! – https://www.qcshub.org • Qiskit – open source https://qiskit.org/ • IBM Quantum – https://www.ibm.com/quantum-computing/ • Quantum Apalooza – https://quantumapalooza.com/ • Quantum Algortihms – https://quantumalgorithmzoo.org/

• Lots of books too! For BCS Berkshire Members

Thank you! Find us at qcshub.org