Quantum Technology Overview (Computing and Communications).pdf

《Quantum Technology Overview (Computing and Communications).pdf》由會員分享，可在線閱讀，更多相關《Quantum Technology Overview (Computing and Communications).pdf（47頁珍藏版）》請在三個皮匠報告上搜索。

1、Cliff Grossner,Ph.D.Bijan NowrooziYasser Omar,Ph.D.Vicente Martin,Ph.D.Quantum Technology Overview(Computing and Communications)Quantum Technology a Natural Extension for OCPCPUGPUQuantumComputationElectricalOpticalQuantumCommunicationHybrid Data CenterWhat is a Quantum Computer?What are the challen

2、ges of Quantum Computation?Introduction to Quantum ComputationYasser OmarYasser Omar,President,PQI Portuguese Quantum Institute,and Professor,IST University of LisbonIntroduction to Quantum ComputationSPECIAL FOCUS:QUANTUMA machine that processes Quantum Information i.e.,it exploits quantum superpos

3、ition and quantum entanglement,properties of Nature accessible(only)at the atomic and sub-atomic scale,to compute.Namely to potentially compute significantly faster than classical machines.And,possibly,to perform significantly more energy-efficient computation.What is a Quantum Computer?How to scale

4、 quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semiconductors,photonics,Quantum Computation:ChallengesHow to scale quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semiconductors,photonics,How to network quantum computers?QLANs as a strategy for scalability,Qua

5、ntum Internet(e.g.for multiparty secure computation),Quantum Computation:ChallengesHow to scale quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semiconductors,photonics,How to network quantum computers?QLANs as a strategy for scalability,Quantum Internet(e.g.for multiparty secure

6、computation),What can we do with quantum computers/where can we expect speedups?Quantum simulation:solving problems from Quantum Physics and Quantum Chemistry.Optimisation problems,geometric data,machine learning,differential equations,cryptoanalysis,Your name here!Quantum speedups are hard to find!

7、Quantum Computation:ChallengesExample of European initiative:How to scale quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semiconductors,photonics,How to network quantum computers?QLANs as a strategy for scalability,Quantum Internet(e.g.for multiparty secure computation),What can

8、we do with quantum computers/where can we expect speedups?Quantum simulation:solving problems from Quantum Physics and Quantum Chemistry.Optimisation problems,geometric data,machine learning,differential equations,cryptoanalysis,Your name here!Quantum speedups are hard to find!How to develop hybrid

9、quantum-classical computing?Quantum Computation:ChallengesHow to scale quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semiconductors,photonics,How to network quantum computers?QLANs as a strategy for scalability,Quantum Internet(e.g.for multiparty secure computation),What can we

10、do with quantum computers/where can we expect speedups?Quantum simulation:solving problems from Quantum Physics and Quantum Chemistry.Optimisation problems,geometric data,machine learning,differential equations,cryptoanalysis,Your name here!Quantum speedups are hard to find!How to develop hybrid qua

11、ntum-classical computing?Deployment,integration,middleware,access?Business cases?Standards?Quantum Computation:ChallengesHow to scale quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semiconductors,photonics,How to network quantum computers?QLANs as a strategy for scalability,Quant

12、um Internet(e.g.for multiparty secure computation),What can we do with quantum computers/where can we expect speedups?Quantum simulation:solving problems from Quantum Physics and Quantum Chemistry.Optimisation problems,geometric data,machine learning,differential equations,cryptoanalysis,Your name h

13、ere!Quantum speedups are hard to find!How to develop hybrid quantum-classical computing?Deployment,integration,middleware,access?Business cases?Standards?Can we have energetic advantage?Quantum Computation:ChallengesHow to scale quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semi

14、conductors,photonics,How to network quantum computers?QLANs as a strategy for scalability,Quantum Internet(e.g.for multiparty secure computation),What can we do with quantum computers/where can we expect speedups?Quantum simulation:solving problems from Quantum Physics and Quantum Chemistry.Optimisa

15、tion problems,geometric data,machine learning,differential equations,cryptoanalysis,Your name here!Quantum speedups are hard to find!How to develop hybrid quantum-classical computing?Deployment,integration,middleware,access?Business cases?Standards?Can we have energetic advantage?Quantum Computation

16、:ChallengesHow to develop hybrid quantum-classical algorithms?Why?-Because quantum processors are limited?-Because classical supercomputers are limited?For which applications?How do hybrid algorithms perform?How to implement them?Hybridising Quantum and Classical ComputingHow to develop hybrid quant

17、um-classical algorithms?Why?-Because quantum processors are limited?-Because classical supercomputers are limited?For which applications?How do hybrid algorithms perform?How to implement them?Hybridising Quantum and Classical ComputingTwo methods to break down a quantum algorithmPartitions the input

18、 into smaller sub-problemsM.Mura,D.Magano,Y.Omar,Physical Review A 109,022412(2024).U requires individually less coherent queries,but collective set yieldsthe same as a single run of U.How to scale quantum hardware?Platforms:cold atoms,trapped ions,superconductors,semiconductors,photonics,How to net

19、work quantum computers?QLANs as a strategy for scalability,Quantum Internet(e.g.for multiparty secure computation),What can we do with quantum computers/where can we expect speedups?Quantum simulation:solving problems from Quantum Physics and Quantum Chemistry.Optimisation problems,geometric data,ma

20、chine learning,differential equations,cryptoanalysis,Your name here!Quantum speedups are hard to find!How to develop hybrid quantum-classical computing?Deployment,integration,middleware,access?Business cases?Standards?Can we have energetic advantage?Quantum Computation:ChallengesTowards Energetic Qu

21、antum Advantage?Estimated energy cost E(J)for increasing input size N of implementing the Quantum Fourier Transform on a trapped-ion quantum computervs.the discrete Fourier Transform on the supercomputers Frontier and Henri.F.Gis,M.Pezzutto,Y.Omar,to appear soon(2024).Quantum Computing Call to Actio

22、n We need:Better,scalable,quantum hardware(including QRAMs).Solutions to exploit networked quantum hardware.More quantum software:new quantum algorithms/applications with quantum speedup Including strategies and methods to harness Hybrid Quantum-Classical Computing.More quantum middleware,standards,

23、integration solutions.To manage expectations/the quantum hype.To explore novel directions,such as potential quantum advantage in energy consumption in computation,both classical and quantum.Introduction to Quantum CommunicationsVicente MartinU.Politcnica de Madrid Vicente.Martinupm.es Vicente Martin

24、,Professor U.Politcnica de Madrid Vicente.Martinupm.es Introduction to Quantum CommunicationSPECIAL FOCUS:QUANTUMOutline22 Its all about qubits:Quantum Communications Ingredients What the qubits bring to communications:the quantum properties that we use.The difficulties of Quantum communications in

25、networks Some sample networks1)Quantum channel.To send the qubits around2)Qubit preparation and analysis.1)Qubit detectors.2)Qubit emitters.3)Quantum conversion(flying qubits to stationary qubits:matter-photon interface)4)Moderate quantum processing capabilities.5)Quantum memories.1)-2)=Quantum cryp

26、to as we know it(Mostly QKD)+3)-5)=Full quantum communications Including unlimited distance QKD.i.e.Quantum repeaters.Quantum Communications IngredientsNote:some recent schemes without quantum memories,all photonic,henceno need for flying-stationary conversion.It does not reveal allReading the state

27、 of a qubit(measurement):(2+2=1,measurement done in the|0,|1 basis.Note:measurement modifies the state of the qubit.We do not have access to or No-cloning:an unknown quantum state cannot be copied.A Qubit is shy|0+|1|0|1Prob.2Prob.2measurement(Wooters,Zurek 1982.Why took more than 50 years to realiz

28、e this?)Actually,also a consequence of the Linearity of the Schroedinger eq.Computational basisIt does not reveal allReading the state of a qubit(measurement):(2+2=1,measurement done in the|0,|1 basis.Note:measurement modifies the state of the qubit.We do not have access to or No-cloning:an unknown

29、quantum state cannot be copied.A Qubit is shy|0+|1|0|1Prob.2Prob.2measurement(Wooters,Zurek 1982.Why took more than 50 years to realize this?)Actually,also a consequence of the Linearity of the Schroedinger eq.Computational basisWith this we can do quantum cryptography.Actually QKD:Quantum Key distr

30、ibution This is the most advanced Q.Comm.technologyNot to say a little bit strange:For example,the state:Is an entangled state:It is not separable as it cannot be described as the direct product of the states of two single qubit states.Non-local!.Non-classical correlations!E.g.:If the first qubit is

31、 measured and 0 is obtained,then the second qubit is,with certainty,also a 0 even if it is in the other side of the Galaxy-Bell inequalitiesThe famous Bohr-Einstein discussion about quantum mechanicsQubits are funny.=12 00+11R.Doisneau2-qubits stateNot to say a little bit strange:For example,the sta

32、te:Is an entangled state:It is not separable as it cannot be described as the direct product of the states of two single qubit states.Non-local!.Non-classical correlations!E.g.:If the first qubit is measured and 0 is obtained,then the second qubit is,with certainty,also a 0 even if it is in the othe

33、r side of the Galaxy-Bell inequalitiesThe famous Bohr-Einstein discussion about quantum mechanicsQubits are funny.=12 00+11R.Doisneau2-qubits stateWith this we can do quantum repeaters.Unlimited distance quantum communications.Creation of quantum correlations among any two points.i.e.the“full power”

34、of quantum communications This is more difficult to achieveQuantum Cryptography.(actually we are talking Quantum Key Distribution:The“low”hanging fruit)Ingredients:A qubit emitter(think photons):Alice.Can prepare qubits in different states and basis.A qubit receiver:BobCan measure qubits in differen

35、t basisA quantum channel(able to transport the qubits from Alice to Bob)A classical channel(public but authentic)and the spy (Eve)Quantum ChannelClassical ChannelNote:this is the most typical set-up.It is called“prepare and measure”.It is not the only one.Quantum Cryptography.(actually we are talkin

36、g Quantum Key Distribution:The“low”hanging fruit)Ingredients:A qubit emitter(think photons):Alice.Can prepare qubits in different states and basis.A qubit receiver:BobCan measure qubits in different basisA quantum channel(able to transport the qubits from Alice to Bob)A classical channel(public but

37、authentic)and the spy (Eve)Quantum ChannelClassical ChannelNote:this is the most typical set-up.It is called“prepare and measure”.It is not the only one.Creation of a symmetric key between the two ends of the quantum channel.Secret:we can bound the information leaked to the external world as much as

38、 we want.Demonstrable:ITS.No computational assumptions required.Ingredients:A qubit emitter(think photons):Alice.Can prepare qubits in different states and basis.A qubit receiver:BobCan measure qubits in different basisA quantum channel(able to transport the qubits from Alice to Bob)A classical chan

39、nel(public but authentic)and the spy (Eve)Quantum Cryptography in networksLimited reach,point to point.Comm.laserSingle photon(not to scale)extremely weak signals.Noise in the fibre:RamanRaman backscattering of a signal at 1549 nm DOI:10.1063/1.1842862Single Photon(approx.scale)Difficult to detect.A

40、bsorptions Masked by the noiseQuantum Communications and Networks Why is it difficult?150 nm0 km109photons/sec.15 km5 108150 km106300 km1000600 km1 p per 20 min.900 km 1 p per36 yearsLosses in fibre 0.2 dB/km=0.20.8 nm(DWDM)=320 nm(CWDM)Limited reach,point to point.Comm.laserSingle photon(not to sca

41、le)extremely weak signals.Noise in the fibre:RamanRaman backscattering of a signal at 1549 nm DOI:10.1063/1.1842862Single Photon(approx.scale)Difficult to detect.Absorptions Masked by the noiseQuantum Communications and Networks Why is it difficult?150 nm0 km109photons/sec.15 km5 108150 km106300 km1

42、000600 km1 p per 20 min.900 km 1 p per36 yearsLosses in fibre 0.2 dB/km=0.20.8 nm(DWDM)=320 nm(CWDM)(with current technology)Trusted nodes/satellites for long distance*Typical:Today experiments with new protocols-60 dB,600 kmLimited reach,point to point.Comm.laserSingle photon(not to scale)extremely

43、 weak signals.Noise in the fibre:RamanRaman backscattering of a signal at 1549 nm DOI:10.1063/1.1842862Single Photon(approx.scale)Difficult to detect.Absorptions Masked by the noiseQuantum Communications and Networks Why is it difficult?150 nm0 km109photons/sec.15 km5 108150 km106300 km1000600 km1 p

44、 per 20 min.900 km 1 p per36 yearsLosses in fibre 0.2 dB/km=0.20.8 nm(DWDM)=320 nm(CWDM)DWDM up to 160 CWDM18 channels1270-1610 nmSep:20 nmCWDM.DWDM1528.77-1536.86 nm80 channels(50GHz grid.Sep:0.4 nm)Lambda(nm)Life in the Optical FibreLimited reach,point to point.Comm.laserSingle photon(not to scale

45、)extremely weak signals.Noise in the fibre:RamanRaman backscattering of a signal at 1549 nm DOI:10.1063/1.1842862Single Photon(approx.scale)Difficult to detect.Absorptions Masked by the noiseQuantum Communications and Networks Why is it difficult?150 nm0 km109 photons/sec.15 km5 108150 km106300 km10

46、00600 km1 p per 20 min.900 km 1 p per36 yearsLosses in fibre 0.2 dB/km=0.20.8 nm(DWDM)=320 nm(CWDM)Quantum/classical co-propagationIssues(not sharing the infrastructure Expensive!)Limited reach,point to point.Comm.laserSingle photon(not to scale)extremely weak signals.Noise in the fibre:RamanRaman b

47、ackscattering of a signal at 1549 nm DOI:10.1063/1.1842862Single Photon(approx.scale)Difficult to detect.Absorptions Masked by the noiseQuantum Communications and Networks Why is it difficult?150 nm0 km109photons/sec.15 km5 108150 km106300 km1000600 km1 p per 20 min.900 km 1 p per36 yearsLosses in f

48、ibre 0.2 dB/km=0.20.8 nm(DWDM)=320 nm(CWDM)Limited reach,point to point.Comm.laserSingle photon(not to scale)extremely weak signals.Noise in the fibre:RamanRaman backscattering of a signal at 1549 nm DOI:10.1063/1.1842862Single Photon(approx.scale)Difficult to detect.Absorptions Masked by the noiseQ

49、uantum Communications and Networks Why is it difficult?150 nm=0.20.8 nm(DWDM)=320 nm(CWDM)R.DoisneauQuantum Does not play well with(classical)networks.What to do?An ad hoc,parallel network just for the quantum channel Check arXiv:arXiv:0804.0122 An“extreme ad hoc”a network highly tuned for the quant

50、um channel and associated classical channels.Check arXiv:arXiv:1309.3923 Pulling all stops:Fully integrated quantum/classical network.Check arXiv:arXiv:2311.12791v2Global Quantum Comms.Scenario 10 years Framework.First Calls 2018 1000 M All Quantum Tech.10 years Framework First Calls 2022.2000 M A P

51、an-European Q.NetworkQuantum communications InfrastructureEuroQCISimilar programs elsewhere.Japan,China,Australia,S.Korea,Russia also have nation-wide quantum programsTesting and evaluation infrastructure for the European Quantum Communication Infrastructure(EuroQCI)initiative-CNECT/2023/OP/0032Chin

52、aGlobal Quantum Comms.Scenario 10 years Framework.First Calls 2018 1000 M All Quantum Tech.10 years Framework First Calls 2022.2000 M A Pan-European Q.NetworkQuantum communications InfrastructureEuroQCISimilar programs elsewhere.Japan,China,Australia,S.Korea,Russia also have nation-wide quantum prog

53、ramsTesting and evaluation infrastructure for the European Quantum Communication Infrastructure(EuroQCI)initiative-CNECT/2023/OP/0032ChinaMake Quantum Communications Normal Communications.(and deliver the goods)J.P.Brito1,R.Brito1,R.Vicente1,P.Salas1,L.Ortz1,J.Faba1,J.Saez-Buruaga1,J.L Rosales1,A.J.

54、Sebastian-Lombraa1,M.Garca1,D.Rincn3,F.Prez 3,C.Snchez3,J.M.Rivas-Moscoso2,A.Pastor2,J.Morales2,R.Cant2,J.Folgueira2,D.R.Lopez2,V.Martn1EU HE Grant No 101114043Digital Europe ProgrammeGrant 101091638QuantERA II ProgrammeEU H2020 Grant 101017733EU NextGenerationEU(PRTR-C17.I1)and Comunidad de Madrid.

55、Acciones Complementarias.Madrid Quantum.1Center for Computational Simulation and ETSI Informticos,Universidad Politcnica de Madrid 28660 Madrid,Spain2Telefnica Investigacion y Desarrollo,Ronda de la Comunicacion s/n 28050 Madrid.Spain3IMDEA Software/RedIMadrid,28660 Madrid.SpainQUBIP Horizon Europe

56、grant 101119746.PQREACT Horizon grant 101119547Thank you!A panoplia of Quantum NetworksFull metro network:CWDM core+GPON accessMadrid SDN QKD Network(actual layout)200920182014From prototypes in telecommunicationstestbed installations to field deployments in production facilities(Telefnica).First SD

57、N Quantumnetwork deployedin the field in the world.“The Engineering of a SDN Quantum Key Distribution Network”IEEE Comms.Mag.July 2019,Specialnumber“The Future of Internet”doi:10.1109/MCOM.2019.1800763;http:/arxiv.org/abs/1907.00174arXiv:1309.3923arXiv:1907.00174arXiv:1006.1858MadQCI-2023 Largest Qu

58、antum network in Europe to date.26 QKD modulesHeterogeneous.Switched.Managed by UPM SW stack(Coord.)arXiv:2311.12791v2Global view of the SDQKD NetworkQKD SystemLKMSSDN AgentConfigurationInformationAPPSKeysSDN-QKDNodeSDN-QKDNodeSDN-QKDNodeSDN-QKDNodeSDN-QKDNodeSDN-QKDNodeSDN-QKDNodeSDN-QKDNodeSDN Con

59、trollerThe SDN controller manages theRequirements of the quantum and Classical devices to optimize the network.45Key points Dynamical connections Standards/toolchain Integrated in a classical network Part of a security ecosystemETSI ISG-QKD 015ETSI ISG-QKD 004/014MadQCI 202326 Q devices(emitters+rec

60、eivers).Earlysystems:since 3 years.Last systems:6 months.Three different QKD manufacturers+experimental devices.Standard optical transport equipment.Standard encryptors at Level 1 and Level 2(Level 3,IPsec in SW)Important:A real world network(Quantum+classical communications)Shared quantum and Class

61、ical infrastructure,includingoptical fibre.CV+DV systems on the same Fibre.Twoconnected operators.Several(quantum and Classical,QKD&encrypt.)manufacturers.Single SDN control.Two networks.Border node:two domains.Standards.End toend key delivery whatever the manufacturer chain.Heterogeneus in the quan

62、tum and classical part.Real use-cases.85K metricsPQC+QKD testbed.Key point:Real worlddemonstration of theintegration of quantum comms.in telecomand securityecosystems.arXiv:2311.12791v2MadQCI 202326 Q devices(emitters+receivers).Earlysystems:since 3 years.Last systems:6 months.Three different QKD ma

63、nufacturers+experimental devices.Standard optical transport equipment.Standard encryptors at Level 1 and Level 2(Level 3,IPsec in SW)Important:A real world networkShared quantum and Classical infrastructure,includingoptical fibre.CV+DV systems on the same Fibre.Twoconnected operators.Several(quantum

64、 and Classical,QKD&encrypt.)manufacturers.Single SDN control.Two networks.Border node:two domains.Standards.End toend key delivery whatever the manufacturer chain.Heterogeneus in the quantum and classical part.Real use-cases.85K metricsPQC+QKD testbed.Key point:Real worlddemonstration of theintegration of quantum comms.in telecomand securityecosystems.arXiv:2311.12791v2Make Quantum Communications Normal Communications.(and deliver the goods)