quantum computing on a scalable superconducting qubit platform · quantum computing first...
TRANSCRIPT
Stefan FilippQuantum Technology, IBM Research – Zurich
Quantum Team at IBM Research Yorktown Heights
Quantum computing
on a scalable superconducting qubit platform
Brasilien
T.J WatsonAlmaden
Austin Irland
Zürich
Afrika Indien
Haifa
China Japan
Australien
12 labs with over 3,000 researchers around the world3 labs with research in quantum information processing
Copper SOI StrainedSilicon
Dual Core Immersion SiGe High-k eDRAM 3D ChipStacking
AirgapChemically Amplified Resists
Quantum
IBM Research
IBM Research – Zurich Established in 1956, ~ 420 employees
Science & technology, cognitive computing & industry solutions, cloud & computing infrastructure
IBM’s most advanced nanotechnology fabrication (collaboration with ETH, EMPA) and IBM’s world-class noise free labs
Proven track record in science & technology –including 2 Nobel prizes (4 Laureates), 1 Kavli prize
Majority of researchers are Europeans(22 EU countries represented)
Longstanding collaborations with universities, research organizations and industries in EU & worldwide
Investments in EU’s research infrastructure and education
Spin-off and start-up companies
Binnig & Rohrer Nanotechnology Center(BRNC, opened 2011)
3D / hybrid
neuromorphic (cognitive)
quantum computing
First integrated circuit
Size ~1cm2
2 Transistors
Moore’s Law is Born
Intel 4004
2,300 transistors
IBM P8 Processor ~ 650 mm2
22 nm feature size, 16 cores
> 4.2 Billion Transistors
1958 1971 2014
Alternative (co-existing)architectures:
Further issues: power density on chip, speed, interconnects, cost,…
Why Quantum? Why now?
© 2016 International Business Machines Corporation
Solving computational problems requires physical resources (time, memory, and space).
“easy” problems (polynomial efficient) “hard” problems (exponential intractable)
• multiplying numbers• word processing• Sending emails
• Algebraic and Number Theoretic Algorithms (factoring, hidden subgroup)• Combinatorial optimization (traveling salesman)• Machine learning • simulating quantum mechanics
There are problems that are believed to be hard (never) for classical computers to solve.
http://xkcd.com/759/
http://xkcd.com/399/
Conventional/Classical Computing
© 2016 International Business Machines Corporation
Exponential speed-up:A task taking 2100 seconds (1025 days) on a classical computer might take 100 seconds on a quantum computer
The problem ofmultiplication vs factoring:
937 x 947 = N (easy)887339 = p x q (harder)
Modulus (1024 bits):de b7 26 43 a6 99 85 cd 38 a7 15 09 b9 cf 0f c9 c3 55 8c 88 ee 8c 8d 28 27 24 4b 2a 5e a0 d8 16 fa 61 18 4b cf 6d 60 80 d3 35 40 32 72 c0 8f 12 d8 e5 4e 8f b9 b2 f6 d9 15 5e 5a 86 31 a3 ba 86 aa 6b c8 d9 71 8c cc cd 27 13 1e 9d 42 5d 38 f6 a7 ac ef fa 62 f3 18 81 d4 24 46 7f 01 77 7c c6 2a 89 14 99 bb 98 39 1d a8 19 fb 39 00 44 7d 1b 94 6a 78 2d 69 ad c0 7a 2c fa d0 da 20 12 98 d3
1024bit public key:
= p × q
→ just short of impossibleShor’s algorithm jumpstarted the interest in quantum computing
Classical Record: 230 digits
exp(𝐶 𝑏1/3)
𝐶 𝑏3
Example: Shor’s Algorithms
© 2016 International Business Machines Corporation
How much memory is needed to store a quantum state?
How much time does it take to calculate dynamics of a quantum system?
# qubits quantum state coefficients # bytes timescale
1 𝑎 0 + 𝑏|1⟩ 21 = 2 16 Bytes
2 𝑎 00 + 𝑏 01 + 𝑐 10 + 𝑑|11⟩ 22 = 4 32 Bytes Nanoseconds
8 28 = 256 2kB Microseconds on watch
16 … 216 = 65′536 256 kB Milliseconds on smartphone
32 … ~ 4 billion 256 GB Seconds on laptop
64 …~ information in
internet74 EB
(74 million GB)Years on supercomputer
256 …~ # of atoms in
universe… never
classicalq
uan
tum
The Quantum Advantage – Simulation of physical systems
© M. Troyer
© 2016 International Business Machines Corporation
Goal: Build computers based on quantum physics to solve problems that are otherwise intractable
Develop “Hardware-efficient” apps − Chemical configurations− Optimization
No full error correction available
Demonstrate “quantum supremacy”
5-8 qubits 16-20 qubits50-100+ qubits 105-106 qubits
Small-scale Medium-scale Large-scale
Research level quantum demonstrations
Verify chemistry and error correction principles
Infrastructure & community building
Known and proven speed-up: Factoring Complete quantum molecular
simulations Speed-up machine learning and
database searching
Roadmap:
Challenges: Continued scalability, control and coherence of large systems, cost…
Grand Challenge: Quantum Computing
© 2016 International Business Machines Corporation
microwave resonator: read-out of qubit states
multi-qubit quantum bus
noise filter
(fixed-frequency) superconducting qubit: non-linear Josephson Junction (Inductance)
anharmonic energy spectrum → qubit
nearly dissipationless → T1, T2 ~ 70 µs
Superconducting Qubit Processor
© 2016 International Business Machines Corporation
Required Technologies for Scaling
• Two-qubit gates: improve fidelities
• Coherence/reproducibility:
improved material/fabrication methods
• Cross-talk: microwave mode suppression
(air-bridges through-vias, solder-bump bonds, flip-chip,…)
• 3D integration: address/couple all qubits in 2D array
• Integrated electronics: required for 20+ qubits
• Verification and characterization: identify Hamiltonian
• Software: compilers and high-level programming language
• …
Quintana (2014)
@ IBM Research - Zurich
MIT-LL
[1] Threshold for surface code QEC assuming 30-100ns gate time
[2] J. Martinis et al., PRL 95 210503 (2005)
[3] K. Geerlings et al., APL 192601 (2012)
[4] H. Paik et al., PRL 107, 240501 (2011)
[5] R. Barends et al., APL 99, 113507 (2011)
[6] A. Corcoles et al., APL 99, 181906 (2011)
[7] C. Rigetti et al., PRB 86, 100506 (2012)
[8] M. Reagor et al., arXiv:1302.4408 (2013)
[9] J. Chang et al. APL 103, 012602 (2013)
[10] P. Kumar, et al. arXiv:1604.00877v1, (2016).
Developments to extend coherence times:
• Materials e.g. [2]
• Design and geometries e.g. [3]
• 3D transmon [4]
• IR Shielding [5,6]
• Cold normal metal cavities and cold qubits [7]
• High Q cavities [8]
• Titanium Nitride [9]
• UHV packaging [10]
Remarkable progress over the past decade!
@ IBM Research - Zurich
Qubit Coherence and Lifetimes
UHV sealing and surface preparation
Cross-Resonance entangling gate (ZX) [1,2]• drive qubit A at frequency of B• fixed frequency qubits → long coherence• fixed coupling → qubits close in frequency• strong MW driving → off-resonant interactions• Typical specs: T1,T2 =60 − 80𝜇𝑠, TG > 160ns,
gate fidelities 95%-99% [Sheldon et al., arXiv:1603.04821 (2016)]
2-Qubit Gates
[1] J. M. Chow, et al. PRL 107, 080502 (2011).
[2] S. Sheldon, et al, PRA, 93, 012301 (2016).
𝛿12
Σ12
Gate with tunable bus (XX,YY,ZZ) [3,4]• modulate J at sum and difference frequencies• fixed frequency → long coherence• simultaneously create XX, YY & ZZ interactions• couple more than 2 qubits by frequency addressing
[3] P. Bertet, PRB, 73, 064512 (2006)
[4] D. C. McKay, et al. PR Appl (2016).
@ IBM Research - Zurich
Entanglement generation via tunable couplings
Randomized Benchmarking:
97.2 % gate fidelity; TG = 140 ns [D. McKay et al. PR Applied (2016)]
𝑖𝑆𝑊𝐴𝑃 with 𝐻𝑡𝑐 ∝ 𝑋𝑋 + 𝑌𝑌
Target state: 1/ 2( 01 + 𝑖|10⟩)~ 97% state fidelity; 80ns
2 − 𝑒𝑥𝑐𝑖𝑡𝑎𝑡𝑖𝑜𝑛 with 𝐻𝑡𝑐 ∝ 𝑋𝑋 − 𝑌𝑌
Target state: 1/ 2( 00 + |11⟩)~ 97% state fidelity; 175ns
(work in progress)
Hamiltonian: 𝐻𝑡𝑐 = 𝐻1 +𝐻2 + 𝐽𝑥𝑥𝑋𝑋 + 𝐽𝑦𝑦𝑌𝑌 + 𝐽𝑧𝑧𝑍𝑍
Goal:
• optimizing gate count
for short-depth circuits
• analog quantum simulation
© 2016 International Business Machines Corporation
Goal: Build computers based on quantum physics to solve problems that are otherwise intractable
Develop “Hardware-efficient” apps − Chemical configurations− Optimization
No full error correction available
Demonstrate “quantum supremacy”
5-8 qubits 16-20 qubits50-100+ qubits 105-106 qubits
Small-scale Medium-scale Large-scale
Research level quantum demonstrations
Verify chemistry and error correction principles
Infrastructure & community building
Known and proven speed-up: Factoring Complete quantum molecular
simulations Speed-up machine learning and
database searching
Enable secure cloud computing: quantum analogy to homomorphic encryption
Roadmap:
Challenges: Continued scalability, control and coherence of large systems, cost…
Grand Challenge: Quantum Computing
© 2016 International Business Machines Corporation
Superconducting 5 Qubit Processor in the cloud
• explore quantum computing: run algorithms and experiments
• build a Quantum Community
• discover new applications for this technology
program 5 qubits
2016 IBM introduced a 5 qubit QC on the cloud for public use: http://www.ibm.com/quantumexperience
© 2016 International Business Machines Corporation
• Several professors committing to using IBM Quantum Experience for courses• Undergrad conference at University of Waterloo using IBM QX• ‘Live the Quantum Experience’ student’s event at IBM Zurich - Research
• >36000 users subscribed• 230000 algorithms run
Are you an experiencer, yet?
www.ibm.com/quantumcomputing
Quantum community
© 2016 International Business Machines Corporation
10+ papers based on IBM QX submittedHigh-level interface: ProjectQ (see M. Troyer)
Quantum computer for scientists
© 2016 International Business Machines Corporation
Five-qubit quantum processor
IBM Quantum Experience
Q0
Q1
Q2
Q3
Q4
Five-qubit device parameters
Frequency(GHz)
Anharm.
(MHz)
Readout cavity (GHz)
T1 (μs) T2 (μs) Readout fidelity
RB Error/Clifford
Q0 5.3503 -330.0 6.5251 57.3 66.4 0.971 0.0027
Q1 5.3061 -328.9 6.4760 71.8 90.7 0.972 0.0018
Q2 5.1202 -322.0 6.4295 63.4 53.4 0.972 0.0025
Q3 5.2297 -327.0 6.5743 82.7 66.9 0.952 0.0032
Q4 5.0748 -330.5 6.5243 71.0 72.2 0.961 0.0025
Q0
Q1
Q2
Q3
Q4
Values from Calibration on 2016-07-18_12.05
Two-Qubit CR (ZX90)
RB Error/ Clifford
(±𝟎. 𝟎𝟎𝟐)
CR0-2 (467ns) 0.044
CR1-2 (573ns) 0.035
CR3-2 (547ns) 0.036
CR4-2 (400ns) 0.033
© 2016 International Business Machines Corporation
Physical layer
Logical layer
The Quantum Computing System
[Gambetta, Chow, Steffen, arxiv 1510.04375 (2015)]
© 2016 International Business Machines Corporation
To detect error: encode 1 bit in 2 bits
Quantum: no measurement without disturbance
Solution: use an ancilla qubit to detect the error – but not the qubit state
0 1 0 1
0 0 1 1
0 1 1 0ancilla qubit 0
dataqubits • Ancilla qubit signals parity of data qubits
• ZZ & XX measurement needed to catch phase and bit flip errors
controlled-NOT gates
ZZ-measurement
parity change:
indicates error
0 → |00⟩
1 → |11⟩
|10⟩
|01⟩
|10⟩
|01⟩
Bit-flip error0 1
1 0
Quantum Error Detection
© 2016 International Business Machines Corporation
Fault tolerant quantum computation via the surface code
[Raussendorf, Harrington, PRL (2007);Fowler et al., PRA (2009); Bravvy (1998)]
• logical qubits formed by specific states of data qubits (delocalized)
• measure syndrome qubits→ gives parity information of data qubits (4 data qubits per syndrome)
• correct for errors on actual data qubits
• error threshold: 𝑝0 = 0.7%
© 2016 International Business Machines Corporation
Scaling superconducting qubit system for surface code
3x3 code
d = 39 code
8 syndromes10 buses
2 qubits, 1 bus 3 qubits, 2 buses
4 qubits, 4 buses 8 qubits, 4 buses
5 IBM QX qubits: 1 plaquette
or
[JMC et al. Nat. Comm (2014), Corcoles et al Nat Comm (2015); Takita, et al., arxiv: 1605.01351 (2016)
[D. Riste et al. Nat. Comm 6, 6983, (2015)
J. Kelly et al. Nature, 519,66, (2015)]
See also:
@ Google, Martinis @ Delft, DiCarlo
© 2016 International Business Machines Corporation
Full plaquette experiment
• full plaquette: syndrome qubit coupled to 4 data qubits via bus resonators
• prepare the 4 data qubits in 16 initial states of various parities (basis:| ۧ0 and | ۧ+ )
• measure the weight-four parities via high-fidelity measurement of Q2, the syndrome
Z-parity check X-parity check
Or
ZZZZ
Par
ity
Ch
eck
syndrome qubit
XX
XX
Par
ity
Ch
eck
syndrome qubit
© 2016 International Business Machines Corporation
State(Q0 Q1 Q3 Q4) Q2 P0 Q2 P1
0 0 0 0 0.878 0.1220 0 0 1 0.191 0.8090 0 1 0 0.120 0.8800 0 1 1 0.908 0.0920 1 0 0 0.188 0.8120 1 0 1 0.815 0.1850 1 1 0 0.914 0.0860 1 1 1 0.100 0.9001 0 0 0 0.128 0.8721 0 0 1 0.878 0.1221 0 1 0 0.936 0.0641 0 1 1 0.107 0.8931 1 0 0 0.887 0.1131 1 0 1 0.229 0.7711 1 1 0 0.125 0.8751 1 1 1 0.922 0.078
Example, preparing data qubit state |1001〉
Syndrome state
|0〉
even
|1〉
odd
Mean of correct parity probabilities: 0.872
ZZZZ-full plaquette result
© 2016 International Business Machines Corporation
State(Q0 Q1 Q3 Q4) Q2 P0 Q2 P1
++++ 0.847 0.153+++- 0.202 0.798++-+ 0.117 0.883++-- 0.908 0.092+-++ 0.178 0.822+-+- 0.811 0.189+--+ 0.922 0.078+--- 0.129 0.871-+++ 0.142 0.858-++- 0.883 0.117-+-+ 0.898 0.102-+-- 0.127 0.873--++ 0.918 0.082--+- 0.150 0.850---+ 0.194 0.806---- 0.860 0.140
Example, preparing data qubit state |+---〉
Mean of correct parity probabilities: 0.863Similar work done for 7Q device: Takita, et al., arxiv: 1605.01351 (2016)
|0〉 |1〉
even odd
XXXX-full plaquette result
Syndrome state
17Q: [[9,1,3]] code would demonstrate thesmallest FT logical qubit
in our architecture
3Q: Z parity check [1]
4Q: [[2,0,2]] error detection, entangled
state codewordprotection [2]
7Q: [[4,1,2]] codespace with at least 2 states, show simple logical
operation and fault tolerance measurement
Demonstrated / under test
LOGICAL QUBITS
[1] JMC et al. Nat. Comm (2014), [2] Corcoles et al Nat Comm (2015)
Growing the lattice
11Q: [[6,1,{3,2}]] could allow investigation of
biased noise
INTERESTING DEMO
LOGICAL OPERATIONS
29Q: [[9,1,3]] with transversal H
2014 2015 2016
© 2016 International Business Machines Corporation
Important steps:
Benchmarks of small-scale quantum systems (logical qubits, simple molecular simulations)
Show quantum advantage in medium-size quantum applications(quantum simulation, quantum optimization, gadget approaches)
Develop test beds for universal quantum computers
Develop a quantum software platform
Further develop understanding of quantum information systems (algorithms, complexity classes, error correction codes,…)
Challenges: Continued scalability, control and coherence of large systems, cost, applications…
Quantum ecosystem:
Expand & define quantum education programs
Establish European quantum technology landscape
Perspective
© 2016 International Business Machines Corporation
Control Software
Cryogenics and Control Electronics
System Characterization
Fabrication/3D Integration
Microwave circuit design & Quantization
Numerical High Frequency Simulation
System Simulation
Superconducting Quantum Processor
can be engineered builds on existing technologies challenges in coherence, control
complexity and scaling
Quantum Algorithms
Quantum Engineering – The Eco System – The Opportunity