Quantum algorithm development

Applications in Genoinformatics

Quantum Computer Architecture

Ir. Aritra Sarkar

PhD candidate
Quantum Computer Architecture Lab
Delft University of Technology
Landscape of Q algorithm development

Quantum Computer
Mechanics Science

Quantum Moore’s
Computing Law

Universal
QC
Fault-tolerant Quantum
QC Accelerators

NISQ
PQC

Quantum Quantum
Supremacy Advantage

2
 Room at the bottom

• QM is an upgrade over EM
– What do these new arsenal mean for computing?

lion
wolf
tiger
dog
cat

3
 It from qubit

• Quantum Turing Machine

– Universal Turing Machine for classical computing, physical process/algorithm
• QTM can be simulated on UTM
– Exception to the extended Church-Turing thesis
• QTM cannot be efficiently simulated on UTM left/right finite
unitary
state
mover
transform
machine read/write
– Hamiltonian model, Adiabatic model, Circuit model head

… 0 1 1 0 0 0 1 …

infinite tape qubits finite alphabet

• Superposition, entanglement, interference, measurement, tunneling
– Provable advantage: P ⊂ BPP ⊂ BQP
– Reversible computing, energy-efficient computing

4
 Slowdown of Moore’s law

• # transistors on IC x2 in 18 months
– Core of digital revolution
• New applications
• Economy
– Various enabling factors
• Algorithms
• Deep UV photolithography

• Slowdown?
– Quantum tunneling Human data
– Heat, von Neumann bottleneck Computers are
not getting
processing
Need for
radically
needs are
better at different
growing at an
• Alternate technology? Exponential
rate
Exponential
computing
technologies
rate

5
 Quantum algorithms

• Coherent protocols:
– Shor’s factorisation & discrete log (QFT, QPE)
– HHL linear systems
– Grover search
– … most studied algorithms so far

• Characteristics:
– Exponential or polynomial provable speedup
– Oracular, black-box algorithms (CS vs Engg. perspective)
– QRAM or quantum I/O or state preparation+tomography
– Low noise tolerance and high qubit multiplicity (FTQC)

6
 Engineering challenges

• let’s make qubits!

• errr…. errors!
– Environment isolation
• Cool it down – superconductors - fridges
– Quantum Error Correction!
• Shor’s code, Steane code, Surface code
• Error threshold

• Universal QC to Q Accelerators
• FTQC to NISQ

7
QAccelerator stack

Algorithms

Gate-based circuit

eQASM/QISA

Routing and
mapping

Error-correction

Micro-coded
analog pulses

Cryo-electronics

Simulator Quantum chip

8
 QC applications

what CC can what QC can

compute compute
what CC will handle
in the near-future

what CC can what QC can other

compute efficiently compute efficiently accelerators

what CC can what QC will handle

what QC can
compute now in the near-future
compute now

9
 Superposition + Interference

• Red: GPU
– Multiple paths evaluated in parallel, results for every state required

• Blue: Probabilistic computing (BPP)

– Explore large solution space but requires only the min/max/mean answer

• Green: Quantum computing (BQP)

– Destructive interference of amplitudes (non-solutions can be made to cancel out)
– Generalization of probability theory for complex amplitudes https://www.youtube.com/watch?v=kEry1TaN4-k

parallel evolution parallel evolution

embarrassingly parallel
global/local interactions global/local interactions
no interaction
statistical answer statistical answer
need all answers
A ++ A +

B + B -
A B C D A B C D
+++ ++
C C

++ +
D D

Initial Parallel All Initial Parallel Superposition Central Initial Parallel Superposition Central
Superposition Tendency
Data Transform States Data Transform of Probabilities Tendency Data Transform of Amplitudes

10
 NISQ era
error rate

QEC
Classical Simulation Limit

FTQC

NISQ

number of qubits
 NISQ: Noisy Intermediate-Scale Quantum
https://arxiv.org/abs/1801.00862 - John Preskill, Quantum Computing in the NISQ era and beyond

11
 Workhorses

• Peter Shor estimates 2048-RSA requires ~5k qubits (times 10 2-103 physical qubits) & ~107 gates

• Near-term quantum algorithms

– low depth circuits without extensive QEC

– enough qubits to just store the problem
– still solve useful problems with local constraints

– Adaptable optimization algorithms

• Easy to map to problem
• Genetic algorithms, deep learning

– Potential to exhibit near-term quantum advantage

https://www.bcg.com/en-ca/publications/2018/next-decade-quantum-computing-how-play.aspx

12
 Quantum software 2.0

• Software 1.0 - explicit instructions to the computer written by a programmer

• Software 2.0 - specify desirable behavior of a program and search

13
 VHQCA

• Variational Principle
– Numerator = E0 (smallest eigenvalue) for some parameter
– Denominator = 1 (if no leakage)
– Possible to reach the ground-state energy by finding the right parameters
• Energy proportional to the cost function of application
• Ground state is optimal solution

• Hybrid Quantum Classical approach (HQC)

– QC used for preparing a quantum state 𝜓 𝜃
• The circuit has a certain format (ansatz/stencil) with parameters
• Measure expectation value 𝜓 𝜃 ℋ 𝜓 𝜃
– Finding optimized parameters 𝜃 offloaded to classical co-processor
– Algorithm complexity traded-off for multiple cycles of initialize-evolve-measure

14
 Supremacy – Advantage – Value

• Weak Quantum Value

• improving solution of any problem, using a QC (possibly in combination with classical methods)
so that the results are better than any available purely classical solution

• Quantum Value
• improving solution of a valuable problem, using a QC (possibly in combination with classical
methods) so that the results are better than any available purely classical solution

• Weak Quantum Supremacy

• solution of any problem using a QC, faster or better than any available classical solution

• Quantum Advantage
• solution of a valuable problem using a QC, faster or better than any available classical solution

• Quantum Supremacy
• mathematical proof that any problem has a super-polynomial separation w.r.t. assumptions
(e.g. P ≠ NP) between any possible quantum algorithm and any possible classical algorithm
• the exhibition of the solution of this problem by a QC at a performance (size, speed, or
efficiency) that is infeasible with any available classical computer

• Strong Quantum Advantage

• Q supremacy + Q advantage (e.g. breaking 2048 RSA with Shor’s algorithm)

15
 My quantum journey
Quantum Mechanics and Quantum Computation (Umesh Vazirani) 2012 – Coursera

Classical
System design; assembly-level satellite memory management; etc.
(Scientist @ Indian Space Research Organisation)

Fundamentals of Quantum Information (Leo DiCarlo)

Quantum Communication and Cryptography (Stephanie Wehner, Thomas Vidick) 2016 – QuTech – MSc Computer Engineering
Quantum Computer Architecture lab
Electronics for Quantum Computation (Carmen G. Almudever, Fabio Sebastiano) Department of Quantum & Computer Engineering
Quantum Information Project (David Elkouss)
2017 – QuTech – MSc Thesis (cum laude)
1 Quantum algorithms for pattern matching on genomic sequences
(Koen Bertels, Zaid Al-Ars, Carmen G. Almudever)

2 Quantum-accelerated de novo assembly using QAOA

2018 – QuTech – PhD candidate
(Koen Bertels, Zaid Al-Ars)
3 Quantum-accelerated algorithmic feature learning in DNA sequences

16
 Quantum algorithms for pattern matching on genomic sequences

RG: Reference Genome (3 × 109𝑏𝑝)

90+ read-aligners
1 size doesn’t fit all!

10
11

16

21
22

27
12
13
14
15

17
18
19
20

23
24
25
26

28
29
30
31
0

5
1
2
3
4

6
7
8
9
0
1
2
3
4
unless…
SR: Short Read (50𝑏𝑝)

 Amino-acid Sequencing
 Motif Finding
 Pattern based Trading
https://arxiv.org/abs/1909.05563
QiBAM - An algorithm for DNA read alignment on quantum accelerators  Speech Recognition

17
 Quantum-accelerated de novo assembly using QAOA/QA

 Warehousing
 Traffic Routing
https://arxiv.org/abs/2004.05078
QuASeR - Quantum Accelerated De Novo DNA Sequence Reconstruction  XNA-GAN

18
 Quantum-accelerated algorithmic feature learning in DNA sequences

“Science is beautiful when

it makes simple explanations
of phenomena or connections
between different observations.
Examples include
the double helix in biology, and
the fundamental equations of
 Bioinformatics physics.”
.
 GRN, GWAS, GO, PPI - Stephen Hawking -
 Artificial Intelligence
 Theoretical Physics

https ://complexitycalculator.com/
https ://www.nature.com/articles/s41586-020-2286-9
https://arxiv.org/abs/2006.00987
Quantum Accelerated Estimation of Algorithmic Information

19
 Quantum-accelerated Genoinformatics

12

17

23

28
10
11

13
14
15
16

18
19
20
21
22

24
25
26
27

29
30
31
1

6
0

2
3
4
5

7
8
9

3
0
1
2

4
21

Genoinformatics
Artificial General Intelligence
Artificial Life
Causal Inference
Digital Physics

20
 Research themes

Communications Molecular Sensing and Optimisation and

Foundations
and Cryptography Simulations Metrology Learning

secure transfer
drug discovery, quantum tunneling quantum biology,
of sensitive EEG, MRI, PPI, …
protein folding in nanopore seq. emergence of life
medical data

Q Genome Q Genome Q Genome

Accelerator Sequencing Analysis

Architecture Stack Qubit Chip

21
 Takeaways

• What Quantum: (in past lectures)

• Why Quantum:
– Can QC help in improving computing hardware (to reset the Moore’s law slowdown)?
– Will QC help in improving computing software (to transcend to radical new applications)?
• How Quantum:
– Active research field - many low-hanging fruits, scarce past references
– Open collaborative community (for QAlgo, not QPU)
– Have theoretical backing to your claim
• Divide risk by exploring mix of PQC and coherent algorithms
– Experiment and demonstrate a proof-of-concept
• Test the problem mapping using simulators
• Try it on available cloud-based QPUs (different architectures: QA, gbQC, CVQC)
• When Quantum:
– QAdvantage vs QWinter
– Tipoff point for (societal/industrial/scientific) application

22