0% found this document useful (0 votes)
34 views

Lecture 6 - Quantum Computing Hardware & Superconducting Circuits

The document discusses quantum computing hardware and superconducting circuits. It introduces qubits and describes how real qubits are implemented using atoms, artificial atoms like transmon qubits, and superconducting circuits. Transmon qubits use a Josephson junction shunted by a capacitor to form an anharmonic oscillator, with the lowest two energy levels acting as the qubit states. Controlling and measuring transmon qubits requires understanding their quantum behavior based on circuit quantization and the Josephson junction Hamiltonian.

Uploaded by

hu jack
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
34 views

Lecture 6 - Quantum Computing Hardware & Superconducting Circuits

The document discusses quantum computing hardware and superconducting circuits. It introduces qubits and describes how real qubits are implemented using atoms, artificial atoms like transmon qubits, and superconducting circuits. Transmon qubits use a Josephson junction shunted by a capacitor to form an anharmonic oscillator, with the lowest two energy levels acting as the qubit states. Controlling and measuring transmon qubits requires understanding their quantum behavior based on circuit quantization and the Josephson junction Hamiltonian.

Uploaded by

hu jack
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 45

Quantum Computing

Hardware & Superconducting


Circuits
Olivia Lanes, PhD
North American Lead, IBM Quantum +
Qiskit Community
Big Picture + Game Plan

(1) What is a REAL qubit anyway

(2) Atoms & artificial atoms


(3) DiVincenzo Criteria

(4) Classical circuits


(5) Quantum Circuits and the
Josephson Junction
(6) How to control & measure a
transmon
2
Motivation

3
How do we build a quantum computer?
|𝑒⟩

Well-behaved quantum systems

|𝑔⟩

…that we can initialize into a known state

…with relatively-long coherence times `

...and a universal set of quantum gates

And qubit-specific measurement capability 4


D. DiVincenzo, arXiv (2020)
What is a real qubit?

3
2
Energy

Toy Model Real Life


5
Images from IonQ, IBM, Phys.org
Qubit Flavors

6
Fig. Source: Michel Kurek, Feb 2021.
Quantum Origins Excited
hydrogen
sample Detector
Prism
Slit

Emitted
Wavelengths

7
Image from chemistrygod.com
Spectral lines
Hydrogen 3
2
Helium

1
Neon
Energy is
quantized!!
Sodium

0
Mercury

8
Image NMSU, N. Vogt
Atomic Qubits

Energy
1
~ "𝑟

1
Only a laser with this energy will
affect our qubit!
0

9
Why make an artificial atom?

CJ

LJ

• Can mimic electromagnetic spectrum


• Can create “knobs” to control all elements we care about
• Can leverage semiconductor fab industry

10
Classical electrical circuits

Toolkit:

R C L

• Implements electrical • Stores energy in • Stores energy in


resistance electric field magnetic field

• Described by Kirchoff’s laws

• Flux and charge are continuous variables


11
What is flux/charge
%'
F= 𝐵𝐴cos(θ) 𝑄 = + 𝐼 . 𝑑𝑡
%& F
+Q

-Q
A

LC circuit
Classical Hamiltonian
1 2 1 2
H= Q +
2C 2L
p 12
!0 /2⇡ = 1/2⇡ LC
Linear circuits

C L

K V
𝑄! ϕ!
𝐻= +
2𝐶 2𝐿
𝑝! 𝑘𝑥 !
𝐻= +
Image taken by O.Lanes @ Univ. of Pittsburgh
2𝑚 2
Quantum linear circuits
IBM Quantum

Means it’s a quantum E


variable now
Quantum Hamiltonian

1 2 1 ˆ2 .
Ĥ = Q̂ + .
2C 2L
.
Ĥ = ~!0 (n̂ + 1/2) |3⟩
|2⟩ ℏ𝜔0
|1⟩ ℏ𝜔0
Need unique addressability for use |0⟩ ℏ𝜔0
as a qubit
F
IBM Confidential – Internal use only
Superconductivity
IBM Quantum

• Vanishing electrical resistance. Non-


dissipative

• Cooper pairs: Electrons anti-correlated


in momentum. Attractive interaction

Resistance
mediated by the lattice

𝑘 −𝑘
e e

Images from physicsfeed.com IBM Confidential – Internal use only


Josephson Junctions
Qubit

𝜙
I
S
=
100 nm

𝐼 = 𝐼" 𝑠𝑖𝑛(𝜙) Amplifier

E = 𝐸# 𝑐𝑜𝑠(𝜙)
1um
! !!
𝜙 = , 𝜑# =
"! $% 16
Images taken by O.Lanes @ Univ. of Pittsburgh
Quantum non-linear circuits: Josephson junction
IBM Quantum

SC p i I
1 = ⇢1 e 1

Insulator RN
p i IC
SC 2 = ⇢2 e 2

Al V

Al O33
Al22O -IC At T=0 K

I C RN =
Al 2e
V. Ambegaokar, A. Baratoff, PRL 10,
486-489 (1963)
JAP, 125(16), 165301, 2019 IBM Confidential – Internal use only
Circuit Quantization ϕ# 𝑄#
𝐻= +
2𝐿 2𝐶
Φ 𝑄#
𝐻 = −𝐸! cos +
𝜑" 2𝐶!
CJ
$! % # $! % ' )#
= − +⋯ +
# &" '! &" #+
LJ 1 𝐸, Higher order
=
𝐿, Φ-' perturbation

Approximate JJ
Hamiltonian: ≈ ħω, 𝑎- 𝑎 − λħ𝑎- 𝑎𝑎- 𝑎 + ⋯

For more details on this math, see Zlatko’s lecture


from QGSS #1
Transmon qubit

Josephson junction with shunting capacitor è anharmonic oscillator

Potential energy

|𝑓 ⟩

𝐿, 𝐶, 𝐶 |𝑒⟩

|𝑔⟩

Φ
lowest two levels form qubit
fge ~ 5.0 GHz, fef ~ 4.80 GHz

Koch, et al. Phys. Rev. A (2007)


Anatomy of an IBM transmon
Superconducting Qubit:
§ Josephson Junction as a nonlinear inductor

100 nm
x 100 nm

Superconducting
Microwave Resonators:
§ read-out of qubit states
§ multi-qubit quantum bus
§ filters at qubit freq
IBM Quantum / © 2022 IBM Corporation 20
21
Break

22
Classical Non-Demolition Measurement

Demolition (obviously) Non-demolition (yay) 23


Analogy from Howard Wiseman
Quantum Non-Demolition Measurement

LJ CJ LJ CJ L C

Qubit Qubit Cavity

𝜔)./01 ≠ 𝜔+,2013

Direct observation QND

24
Quantum dispersive measurement

𝐻 "
𝛼 "" $
= = 𝜔! 𝑞 𝑞 + 𝜔# 𝑐 𝑐 − 𝑞 𝑞 − 𝜒𝑞" 𝑞𝑐 " 𝑐
ℏ 2

2𝐷 𝐶𝑎𝑣𝑖𝑡𝑦 3𝐷 𝐶𝑎𝑣𝑖𝑡𝑦

Readout dispersive shift 𝜒


I
amplitude |𝑒⟩ |𝑔⟩ |𝑒⟩
$
𝐼( $
+ 𝑄(
|𝑔⟩
𝑓
𝑓)
𝜃
width 𝜅 𝑛D
Readout 𝜋1
phase 2 Q
𝑄 |𝑒⟩ 𝜃 = 2 tan&'
𝜒1
𝜅
tan&' (1𝐼
( |𝑔⟩
− 𝜋12 𝑓
Qubit measurement, no amplifier

Dilution refrigerator ≈ 10 𝑚𝐾

𝜃
|𝑒⟩

|𝑔⟩

I I
|𝑔⟩ |𝑒⟩

𝑛D 𝜃
𝑛D
Q Q
Qubit measurement, HEMT only

|𝑒⟩

Room temp

|𝑔⟩ HEMT

I I I
|𝑔⟩ |𝑒⟩

𝑛D 𝜃 𝜃
𝑛D
𝐺 𝑛D
Q Q Q
Qubit measurement with preamp

|𝑒⟩

Q-L preamp
|𝑔⟩ HEMT

I I |𝑔⟩ I |𝑒⟩ |𝑔⟩ I |𝑒⟩


|𝑔⟩ |𝑒⟩

𝑛D 𝜃 𝜃 𝜃
𝑛D 𝑛D 𝑛D
Q Q Q Q
• State preparation
• msmt back-action
• error correction
C. Caves 1982 • ……
IBM Quantum
High fidelity single-shot readout

|𝑒⟩
• State of the art fidelities exceed 99%

• Josephson junction based quantum


limited amplifiers enable single-shot
|𝑔⟩ measurement

Quantum jumps
R. Vijay et al, PRL 106, 110502 (2011) IBM Confidential – Internal use only
A theoretical look at gates
IBM Quantum

Pauli Matrices:

Rotations about the


Bloch sphere:
𝑅 𝜃 = exp −𝑖 𝜃>2 𝑛 . 𝜎 = 𝐼 cos 𝜃>2 − 𝑖 𝑛 . 𝜎 𝑠𝑖𝑛 𝜃>2

𝜎 = 𝑋, 𝑌, 𝑍 |0 ⟩

cos 𝜃(2 −𝑖sin 𝜃(2


Example of X rotation: 𝑅! 𝜃 =
−𝑖sin 𝜃(2 cos 𝜃(2

|1⟩
IBM Confidential – Internal use only
Classical vs. quantum gates
Classical Computers
NOR
NAND
A B Output A B Output
0 0 1 0 0 1
1 0 1 0 1 0
0 1 0 1 0 0
1 1 0 1 1 0
Quantum Computers

cNOT Input Output cPhase Input Output iSWAP Input Output


|00⟩ |00⟩ |00⟩ |00⟩ |00⟩ |00⟩
|01⟩ |01⟩ |01⟩ |01⟩ |01⟩ 𝑖 |10⟩
|10⟩ |11⟩ |10⟩ |10⟩ |10⟩ 𝑖 |01⟩
|11⟩ |10⟩ |11⟩ −|10⟩ |11⟩ |11⟩
31
How do we interact with Microwave Generator

superconducting qubits?

Attenuated coax lines


Quantum Processor
IBM Quantum / © 2022 IBM Corporation 32
Generating Shaped Microwave Pulses
IBM Quantum

𝛼 M
𝐻 = 𝜔L 𝑎- M 𝑎-
+ 𝑎- 𝑎- 𝑎- M 𝑎- − 1
2
+Ω 𝑎- M + 𝑎- cos 𝜔N 𝑡 + 𝜙

This term is physically generated by


applying an oscillating voltage
(microwave pulse) at the qubit

See, e.g., McKay et al, Phys. Rev. A 96, 022330 (2017)

Arbitrary waveform generators (which can


produce arbitrary voltages at ~1-2
GSamples/s) are used to produced shaped
pulses at lower frequency which are then
mixed up with an IQ mixer to the qubit t im e
frequency
IBM Confidential – Internal use only
Single Qubit Control

Z Rotations

Axis of rotation in Bloch sphere


depends on phase
Rotations come for free:
Just shift phase of
subsequent pulses

IBM Quantum / © 2022 IBM Corporation 34


Two Qubit Gates

Cross Resonance: ZX Operation

Rotation of Target Qubit depends on


state of Control Qubit

target
Q2
J ~
bus
Q1
Control Qubit Target Qubit control

35
[C Rigetti et al, PRB (2010)] [JM Chow et al, PRL (2011)]
Large Processor Development

36
Development roadmap

IBM Quantum / © 2022 IBM Corporation 37


Putting it all together

Radio-
Room temp frequency
control (RF) cables
electronics

Qubits + cryo
amplifiers

Maika J

38
Quantum Hardware Challenges

Inside the Cryo Dilution Fridge Processor, device


Room temp development
Cryo-CMOS controls
electronics Coherence, junctions, materials
(stable, low-noise, cost)

Better two-qubit gates

Cryo flex lines

Novel qubit couplers

w
Amplifiers, Attenuators, Isolators

39
Coherent errors
Q2
1 0 0 0 0.995 0.1𝑗 0 0
Q1
0 1 0 0 −0.1𝑗 0.995 0 0
𝑈!"#$ = 𝑈%&' =
0 0 0 1 0 0 −0.1 0.995
0 0 1 0 0 0 0.995 −0.1

Q3
-3
x 10
-0.8

-1
• Always-on interactions
Signal (a.u.)
-1.2

-1.4
• Spectators (Quantum cross-talk)
-1.6 • Frequency collisions
-1.8
0 5 10 15 20
Number of Repetitions

• Coherent over/under rotations


• Calibration of pulses by error amplification
• Incomplete understanding of drive Hamiltonian
Incoherent errors
IBM Quantum

• Qubit decoherence (loss of quantum information)


–T1: relaxation time (decay from |1> to |0>)
–Tϕ: dephasing time (randomization of ϕ)
–T2: overall decoherence time (both T1 and Tϕ)

1 1 1
= +
𝑇# 2𝑇: 𝑇;

IBM Confidential – Internal use only


IBM Quantum
Superconducting qubits coherence timeline
10
-2 – Understand charge noise e.g. [1]
best – 3D transmon [5]
-3
10 repeatable – IR Shielding [6,7],
– Cold cavities & cold qubits [8]
10 -4
– High Q cavities [9]
10 -5 – Materials e.g. [2,10]
– Design and geometries [4,10]
(7s)
TT2 (s)

-6
10
– Microwave environment [3]
2

-7
10 [1] Koch et. al. PRA 76, 04319 (2007)
[2] J. Martinis et al., PRL 95 210503 (2005)
-8 [3] Houck et. al. PRL 101, 080502 (2008)
10 [4] K. Geerlings et al., APL 100, 192601 (2012)
[5] H. Paik et al., PRL 107, 240501 (2011)
-9 [6] R. Barends et al., APL 99, 113507 (2011)
10 [7] A. Corcoles et al., APL 99, 181906 (2011)
[8] C. Rigetti et al., PRB 86, 100506 (2012)
10
-10 [9] M. Reagor et al., APL 102, 192604 (2013)
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 [10] J. Chang et al. APL 103, 012602 (2013)
Year

IBM Confidential – Internal use only


Sources of Leakage • Bandwidth of fast
pulse excites e-f
transitions. • Readout signal can excite
multiphoton transitions to
states well above |f>
• Single qubit gates

• Readout

Sank, PRL 2016.


• Two qubit gates

• Reset

Verney, PRApp (2019)


• Depends on implementation

• CR requires very strong drive tones

• We know to avoid certain collision frequencies


to avoid leakage
43
• Ongoing research
Summary & Take-Aways

• You now know what a real superconducting


qubit looks like, and the components that
make it

• A little something about classical circuits and


quantum circuits

• Why Josephson Junctions are *key*

• How to measure and control a qubit

• Key challenges in the field of hardware


research

44
Thank you

You might also like