Professional Documents
Culture Documents
Quantum Information Science
Quantum Information Science
Q
uantum computing is the use of engineered devices constrained to 50–100 qubits. The million-
quantum systems to perform computation, qubit systems needed for full-scale quantum opera-
meaning physical systems comprised of qua- tions (such as Shor’s factoring algorithm and Grover’s
ntum objects (atoms, ions, and photons) manipulated search algorithm) are not immediately immanent.
through configurations of logic gates. “Quantum” Technical breakthroughs are needed to deliver the
refers to the scale of atoms (nanometers 10 9), ions quantum error correction required to progress from
and photons (picometers 10 12), and subatomic par- NISQ devices to fully fault-tolerant quantum comput-
ticles (femtometers 10 15). The general expectation in ing (FTQC).1 However, progress in many areas of
quantum computing is a quadratic speedup in the foundational physics is enabling new technologies
computation of certain kinds of equations. Algorithms that translate to practical use in quantum computing,
are being rewritten to take advantage of the quantum for example, a black-hole-on-a-chip formulation and
speedup in processing linear algebra routines, Fourier experimental setups for testing quantum gravity in
transforms, and other optimization tasks. the laboratory.2
The development of quantum technologies is a
worldwide effort spanning several hardware platforms
that are currently available via cloud services (e.g.,
Superposition and Entanglement
Superposition and entanglement are two counterin-
IBM Q 27-qubit, IonQ 32-qubit, and Rigetti 19Q Acorn
tuitive features that are harnessed in quantum com-
systems) and that have been announced in research
putation. Superposition is the quantum property of
efforts (Google’s 53-qubit Sycamore quantum proces-
an unobserved particle existing in all possible states
sor). Contemporary quantum systems are nonerror-
simultaneously, until collapsing to only one when
corrected noisy intermediate-scale quantum (NISQ)
measured. Entanglement is the quantum property of
physical attributes being correlated between par-
ticles (position, momentum, polarization, and spin),
This work is licensed under a Creative Commons Attribution 4.0 even when separated by distance (if one particle is
License. For more information, see https://creativecommons. in a spin-up state, the other is in a spin-down state;
org/licenses/by/4.0/
Digital Object Identifier 10.1109/MIC.2021.3132591 a “heads–tails” relationship). The scalability advan-
Date of publication 3 December 2021; date of current version tage of quantum computing is seen in the superpo-
28 January 2022. sition-based modeling of data as the quantum
January/February 2022 Published by the IEEE Computer Society IEEE Internet Computing 7
QUANTUM AND POST-MOORE'S LAW COMPUTING
information representation of all possible system consistent with the macroscopic properties of a ther-
states simultaneously. modynamic system (volume, pressure, and tempera-
ture). Entropy is central to information science as the
minimum number of bits (qubits) needed to communi-
QUANTUM CRYPTOGRAPHY
cate a message, and for measuring the interrelated-
One of the most immediate high-profile applications
ness of subsystems within a system.
of quantum computing is cryptography. The “Y2K of
Entropic uncertainty is a stronger more-comput-
crypto” problem is that practical quantum computing
able version of the Heisenberg uncertainty principle
will break the security of nearly all modern public-key
that concretizes the tradeoff between conjugate var-
cryptographic systems (based on SHA-256 and related
iables such as position–momentum and energy–time
hashing algorithms).3 Postquantum cryptographic
(frequency); the more that is known about a par-
algorithms that are resistant to quantum computer-
ticle’s position, the less that can be known about its
based attacks will need to be implemented in a world-
momentum. Entropic uncertainty constitutes the
wide roll-out effort. The development of quantum-
lower bound of the Heisenberg uncertainty principle
resistant public-key cryptographic standards is under-
and is straightforward to calculate as minimum
way, with the algorithm selection process expected to
entropy (min-entropy) (using entropy instead of stan-
be completed by 2023. In June 2021, the U.S. National
dard deviation).
Institute of Standards and Technology (NIST) anno-
Min-entropy measures the amount of uncertainty
unced that their postquantum cryptography standard-
(entropy) in a system by quantifying the degree of uni-
ization process had entered the third phase, with
formity in the distribution of a random variable (as a
seven third-round finalists and eight alternate candi-
lower bound of the sum of entropies comprised by the
dates being considered for standardization. The main-
temporal and spectral Shannon entropies or (equiva-
stay application for postquantum cryptography is
lently) as the quantum generalization of conditional
quantum key distribution on global networks.
Renyi entropies). The interpretation is that the lower
Quantum-secure algorithms mainly involve a shift
the min-entropy, the higher the certainty of the sys-
to mathematics based on lattices (group theory) as
tem producing a certain outcome (unbreakable cryp-
opposed to factoring (number theory). Some of the first
tography for example). Quantum systems can also be
quantum algorithms developed to take advantage of
upper-bounded using Holevo’s theorem to restrict the
nonclassical properties were Shor’s quantum factoring
amount of information known about a quantum state.
algorithm and Grover’s quantum search algorithm.
Shor’s algorithm is a period-finding function with a
quantum Fourier transform (a classical discrete Fourier QUANTUM MACHINE LEARNING
transform applied to the vector amplitudes of a quan- Quantum machine learning (machine learning applied
tum state), which is exponentially faster than classical in a quantum environment) is emerging as one of the
algorithms (the general number field sieve). More first potential general-purpose applications of near-
recently proposed quantum-secure cryptographic term quantum devices. A key advance is the Born
methods include access based on location instead of machine as the quantum version of the Boltzmann
authorization, with quantum secret sharing localized machine.5 Both “machines” are automated energy
to space-time, and authorized regions replacing autho- functions that evaluate probability output from
rized parties. Time entanglement (entanglement in machine learning algorithms. Classically, the Boltz-
time instead of space) is also possible for crypto- mann machine uses an energy-minimizing probability
graphic key exchange within a short time window. function for sampling (per the Boltzmann distribution
in statistical mechanics). The Born machine interprets
results with the Born rule (a computable quantum
Entropic Uncertainty
mechanical formulation that evaluates the probability
The potential advent of global quantum networks,
density of finding a particle at a given point as being
with end-to-end qubit delivery and entanglement as a
proportional to the square of the magnitude of the
service, suggests that quantum properties such as
particle’s wavefunction at that point).
entropy might be engaged for a sophisticated slate of
applications.4 Entropy has a specific meaning in the
second law of thermodynamics in that the total Born Machine and Quantum GANs
entropy of an isolated system cannot decrease over (QGANs)
time. More generally, entropy quantifies the number of Born machines facilitate a substantial shift into unsu-
microscopic configurations (microstates) that are pervised learning (the classification of unlabeled
scrambling is a measure of how quickly information machine to analyze S&P 500 time-series data for port-
spreads out in a quantum system so that a local mea- folio optimization, and a quantum kernel learning
surement is no longer possible. The information is not method based on RKHS as a data classification
lost and may be recovered later in a different part of method for trade selection.
the system (which has implications for quantum tele-
portation, memory, and cryptography). Quantum infor-
Quantum Walks
mation scrambling has been demonstrated in
Quantum walks are the quantum version of the ran-
Google’s 53-qubit system.
dom walk used in finance to characterize market
behavior on the model of the Brownian motion.12 Clas-
QUANTUM FINANCE sical random walks are based on Markov (stochastic)
processes and a coin flip, and quantum walks too pro-
Quantum finance is the application of physics meth-
ceed on the basis of a coin flip, using a quantum coin-
ods to problems in finance, namely, option pricing,
flip operator such as a Hadamard coin (which flips a
trading strategies, risk management, and portfolio
qubit into a one or zero). Quantum walks also have
optimization. Markets have long been modeled as
more variables to select, including the multidimen-
complex physical phenomena per the principles of
sional graph structure of the walk environment, the
wavefunctions, thermality, dissipation, and Brownian
walk algorithm, the time regime (discrete-continuous),
motion, and now these models are being implemented
and the coin operator.
with real-life quantum hardware instead of simulation.
Whereas random walks occur in the flat spatial
Quantum finance could be one of the first mainstream
structure of a 2-D grid, quantum walks proceed in the
fields to develop in quantum computing as the finan-
multidimensional space of a Bloch sphere, per the
cial industry is typically an early adopter of new tech-
mathematics of lattice graph structures. The lattice
nologies, and currently finds itself in a potential
graphs might be further embedded in a Euclidean
progression from classical to digital (blockchains) to
plane to form regular tessellation patterns (triangular,
quantum methods.
square, and honeycomb lattice graphs).
So far, the main quantum algorithms deployed in
Quantum walks are faster than random walks as
finance include quantum amplitude estimation algo-
they have ballistic propagation through a lattice
rithms, quantum Monte Carlo methods, anharmonic
graph as opposed to classical algorithms that are
oscillators, and quantum kernel learning. Quantum
constrained to diffusive spread. Another factor is
amplitude estimation algorithms are used to estimate
interference effects, which allow the quantum walk
the properties of random distributions such as risk
to penetrate decision-tree structures more quickly.
measures.11 A quantum circuit is defined with a unitary
This enables the primary application of quantum
operator to act on a register of qubits and an estima-
walks, which is finding faster algorithms, for exam-
tion operator to operate on the system based on
ple, in cryptography and search. The faster the algo-
quantum phase estimation (approximating certain
rithm, the safer it is cryptographically (a hacker has
eigenvalues of the estimation operator). Qubits are
less time to infiltrate the system) and the faster the
put into superposition (by applying the Hadamard
search results are delivered (e.g., to identify a trad-
gates) and used to control different powers of the esti-
ing opportunity).
mation operator. The system is evolved (by applying
an inverse quantum Fourier transform) and the qubit
state is measured. The result is an integer that is Quantum Blockchains
mapped to a classical estimation function. Streamlin- Quantum blockchains are proposals for blockchains
ing the phase estimation process requires even fewer (distributed ledger systems) that use quantum methods
gates. Tested on the IBM Q 20-qubit Tokyo system, for postquantum cryptography and the underlying
the method indicates a quadratic speedup compared blockchain protocols. As cryptographic systems, block-
to the convergence rate of classical Monte Carlo chains are acutely exposed to the risk of practical quan-
methods. tum computing and would be likely to upgrade quickly
Other quantum finance projects use quantum once postquantum cryptographic algorithms become
amplitude estimation likewise for bond portfolio standard. Quantum key distribution is an important
value-at-risk assessment per interest rate movement, first-tier application for quantum blockchains, and also
anharmonic oscillators to define a price–energy rela- continuous-time quantum walks, which are faster (and
tionship in Schro € dinger wavefunctions for asset pric- therefore more secure) than classical random walks.
ing and market dynamics, a quantum circuit Born Blockchain protocols might be formulated with optical
networks that rely on entangled Greenberger–Horne– The third level of quantum neuroscience applica-
Zeilinger states, and use anharmonic oscillators for tions is neuroscience physics (the neuroscience inter-
computational verification (zero-knowledge proofs). pretation of foundational physics findings). Using
operators to evolve the system is a key requirement
for the study of quantum systems. Traditionally, quan-
QUANTUM NEUROSCIENCE tum system evolution involves the Schro € dinger and
Quantum neuroscience is the application of quantum Heisenberg dynamics. However, the Heisenberg equa-
information science methods to problems in neurosci- tion of motion is only a general approximation of
ence, particularly wave-based analysis, quantum biol- movement and does not include temperature, and the
ogy state modeling, and neuroscience physics.13 Some Schro € dinger wavefunction is limited to describing
of the first quantum neuroscience applications use pure quantum states as opposed to mixed (subsys-
quantum machine learning to study EEG data with tem) states. Contemporary methods for quantum sys-
quantum neural networks, for example, proposing a tem evolution thus include ladder operators and
quantum circuit to extract 794 features from 21 EEG master equations.
channels to evaluate Parkinson’s disease patients for Ladder operators (creation–annihilation operators)
deep brain stimulation treatment. One research proj- are a standard quantum operator used to raise and
ect has developed quantum algorithms to reconstruct lower quantum system tiers to evolve systems (by act-
medical images from MRI, CT, and PET scans (MRI ing to increase or decrease the eigenvalue of another
images with an inverse Fourier transform, and CT and operator). Ladder operators are a first-line approach,
PET images with an inverse Radon transform and the and more detailed quantum system evolution is pro-
Fourier slice theorem). Another project integrates EEG vided by master equations. The quantum master equa-
and fMRI wave-based data at various spatiotemporal tion is the quantum version of the classical master
scales and dynamics regimes. A key finding is that equation (describing the time evolution of a system as
whereas epileptic seizure can be explained by chaotic a probabilistic combination of states); the simplest
neural dynamics, normal resting states are more com- form is the Lindblad equation (Lindbladian), a quan-
plicated, and are perhaps explained by bifurcation tum Markov model (stochastics with quantum
neural dynamics, in which there is an orbit-based orga- probability).
nizing parameter periodically interrupted by counter- The AdS/Brain theory is introduced as the first
signals to trigger a neural signal. multitier interpretation of the anti-de Sitter space/
The second level of quantum neuroscience appli- conformal field theory (AdS/CFT) correspondence.13
cations is quantum biology state modeling, which The AdS/CFT correspondence is a theory that any
involves superpositioned data (the superposition- physical system with a bulk volume can be described
based modeling of data as the quantum information by a boundary theory in one less dimension. The AdS/
representation of all possible system states simulta- Brain theory is a composite model of neural signaling
neously) and quantum probability.14 Biological sys- in the form of a multitier renormalized brain multiscale
tems constantly interact with their environment and entanglement renormalization ansatz (bMERA) tensor
change their behavior as a result. Models thus include network. The model involves four successive bulk-
updating processes in which the system takes meas- boundary pairs at the scale tiers of brain network, neu-
urements and responds accordingly. ron, synapse, and molecule, all of which are involved in
Quantum biological states are interpreted with neural signaling.
quantum probability, quantum mechanical rules for Other neuroscience physics projects propose AdS/
assigning probabilities that accommodate interfer- Memory (an information storage model based on
ence effects that violate the law of total probability, excited-state energy-tiers), a neuronal gauge theory
and commutativity in conjunction, in classical sys- [in which the system’s global symmetry parameter
tems. A quantum variant of total probability can be (free energy) is rebalanced by gauge fields after signal-
obtained with an interference term for incompatible ing], and a superconducting ordered–disordered phase
observables given by positive operator-valued meas- transition model of neural signaling.
ures, positive measures on a quantum subsystem of
the effect of a measurement performed on the larger
system. The main interpretation of quantum probabil- QUANTUM NETWORKS
ity is with the Born rule, but there can be others. Quan- Global-scale quantum networks are envisioned for
tum Bayesian methods, notably QBism (“cubism”), ultrafast, ultra-secure communication, computation,
incorporate subjective (observer-based) aspects. and sensing. Such networks would likely be photonic,
relying on expertise in global fiber optic communica- basis of operator parameters. For example, winding
tions dating from the 1960s. A full-stack roadmap and unwinding operator size distributions are a means
(analogous to the OSI stack) is proposed for end-to- of guessing wavefunctions relevant to transmission.
end qubit delivery and heralded entanglement.4 Entan- Whereas conventional-size distributions uniformly sum
glement generation is a central requirement for quan- amplitude coefficients to approximate the wavefunc-
tum networks (for quantum key distribution and tion, winding-size distributions use a method in which
quantum teleportation), and various plans outline dis- coefficients in the size basis acquire an imaginary
tilled, swapped, heralded (confirmed), and high-dimen- phase that accelerates the winding and unwinding of
sional entanglement. operator size distribution. The result is teleportation-
by-operator-size, a more efficient information com-
pression protocol. Specific to many-body quantum
Qubits and Qudits
teleportation, operator growth spreading and peaked-
The basic unit of quantum information is the qubit
size teleportation protocols are defined.
(quantum information bit), a two-state quantum-
An advanced method used in quantum teleporta-
mechanical system used for computing, as the quan-
tion is out-of-time-order correlation functions
tum version of the classical bit. However, qubits are
(OTOCs).15 OTOCs are used to evolve a quantum sys-
just one example of the broader class of qudits (quan-
tem backward or forward in time to apply actions and
tum information digits) since a unit of quantum infor-
measure the system. The computation involves calcu-
mation can exist in any number of states. A qubit is a
lating the rate of system divergence by comparing
binary quantum information bit taking the values of 0
how fast two initially commuting operators decay to
and 1 and exists in a superposition of 0 and 1 before
become noncommuting. OTOCs indicate whether a
measurement. A qutrit is a trinary quantum informa-
quantum system is chaotic and its scrambling time.
tion bit taking the values of 0, 1, and 2, and exists in
Chaos describes the evolution of many classical
the 0, 1, and 2 states until measured.
and quantum systems, indicating that seemingly ran-
Qudit-based systems offer greater efficiency than
dom states are in fact governed by deterministic laws
binary systems, for example, the qutrit as a three-state
and sensitivity to initial conditions. Quantum chaotic
unit configures a self-contained nine-state error-cor-
systems are characterized by growth eras of initially
rection code system. Qudits are routinely produced
ballistic growth that then slows and saturates, as
and transmitted in photonic networks, including with
described by the Lyapunov exponent (a quantification
high-dimensional quantum entanglement (entangling
of the butterfly effect). Complex systems are generally
particles in more than one dimension simultaneously,
found to be chaotic and fast-scrambling.
such as polarization, spatial modes, propagation
paths, and time–frequency bins).
The standard errors are a bit flip, a sign flip (the error correction methods cannot be found and plat-
sign of the phase), or both. Basic codes diagnose the forms do not progress beyond hundred-qubit systems.
error, corresponding to the Pauli matrices for control- Sharper critiques are that more foundational quantum
ling qubits in the X, Y, and Z dimensions. The error is algorithms should have been discovered by now, that
expressed as a superposition of basis operations given quantum problem-solving may be limited to certain
by the Pauli matrices. If there is an error, the same kinds of operations (e.g., linear algebra), and that a full-
Pauli operator is applied to act again on the corrupt blown quantum infrastructure (processors, repeaters,
qubit to reverse the error effect. The unitary correc- networks, and memory) is needed to assess the poten-
tion returns the state to the initial state without mea- tial contribution of the technology.
suring the qubit directly. However, despite these challenges, it is not too
The basic form of the quantum error-correcting early to reformulate classical problems for computing
code is the stabilizer code (applied via ancilla), as the in the quantum domain and begin testing on quantum
quantum version of linear codes used classically. In simulators and near-term quantum hardware available
quantum computing, topology-based stabilizer codes via cloud services. At minimum, quantum computing
(the toric code and the surface code) are proposed as provides an alternative platform to ease the burden
a potential method for fault-tolerant error correction. on classical von Neumann architectures in a moment
Particle movement and its correction are interpreted of transitioning to post-Moore’s law computational
in the structure of lattice topologies. Toric codes are solutions. Overall, quantum computing emerges as a
stabilizer codes defined on a 2-D lattice with periodic novel computational platform that provides a more
boundary conditions (thus giving the shape of a torus), capacious architecture with greater scalability and
with stabilizer operators on the spins around each ver- energy-efficiency than current methods of classical
tex and plaquette (face). Surface codes are a more computing and supercomputing and more naturally
generic formulation of topology-based stabilizer corresponds to the 3-D structure of atomic reality.
codes, which are also defined on 2-D spin lattices, and
take various shapes, but are not necessarily toroidal.
Lattice surgery is a method of switching between REFERENCES
codes on the fly. 1. J. Preskill, “Quantum computing 40 years later,” 2021,
There are two kinds of quantum gates, Clifford arXiv:2106.10522.
gates, and non-Clifford gates (based on Clifford alge- 2. A. R. Brown et al., “Quantum gravity in the lab:
bras). Clifford gates are basic quantum gates (Pauli Teleportation by size and traversable wormholes,” 2019,
matrices, and the Hadamard, CNOT, and p/2-phase aXiv:1911.06314v1.
shift gates) that can be simulated efficiently on a clas- 3. W. Barker, M. Souppaya, and W. Newhouse, Migration
sical computer (i.e., in polynomial time). Non-Clifford to Post-Quantum Cryptography. Rockville, MD, USA:
gates are needed for greater logical depth and more NIST Nat. Inst. Standards Technol. Nat. Cybersecurity
complex operations (such as the p/8 gate), and cannot Center Excellence, 2021, pp. 1–15.
be simulated efficiently on a classic computer. 4. A. Dahlberg et al., “A link layer protocol for quantum
Magic state distillation is used to implement non- networks,” in Proc. ACM Special Int. Group Data
Clifford gates, a computationally expensive task of Commun., 2019, pp. 159–173.
consolidating multiple noisy quantum states into a 5. S. Cheng, J. Chen, and L. Wang, “Information
few reliable states. A more efficient alternative is perspective to probabilistic modeling: Boltzmann
gauge fixing. One proposed code based on gauge fix- machines versus Born machines,” Entropy, vol. 20, 2018,
ing is a stabilizer code related to Majorana fermion Art. no. 583.
braiding and color codes (gauge color fixing). Another 6. M. Schuld and N. Killoran, “Quantum machine learning
is a time-based surface code that replicates the 3-D in feature Hilbert spaces,” Phys. Rev. Lett., vol. 122, 2019,
code that performs the non-Clifford gate functions Art. no. 040504.
with three overlapping copies of the surface code 7. E. M. Stoudenmire and D. J. Schwab, “Supervised
interacting locally over a period of time. learning with quantum inspired tensor networks,” Adv.
Neural Info. Proc. Sys., vol. 29, 2017, Art. no. 4799.
8. J. Martyn, G. Vidal, R. C., and S. Leichenauer,
CONCLUSION “Entanglement and tensor networks for supervised
Quantum information science is a field with substantial image classification,” 2020, arXiv:2007.06082v1.
progress and near-term opportunity, but also risks. It 9. Z. Li et al., “Fourier neural operator for parametric
may be too early in the technology cycle if fault-tolerant partial differential equations,” 2021, arXiv:2010.08895v3.
10. X. Mi et al., “Information scrambling in computationally Purdue University. She is the corresponding author of this arti-
complex quantum circuits,” 2021, arXiv:2101.08870v1. cle. Contact her at melanie@blockchainstudies.org.
11. N. Stamatopoulos et al., “Option pricing using quantum
computers,” Quantum, vol. 4, 2020, Art. no. 291.
RENATO P. DOS SANTOS is currently a Researcher in block-
12. V. Kendon, “How to compute using quantum walks,”
chain technologies and a graduate professor with the Lutheran
EPTCS, vol. 315, 2020.
~o Jose
University of Brazil, Sa , Brazil. His research interests
13. M. Swan, R. P. dos Santos, M. A. Lebedev, and F. Witte,
Quantum Computing for the Brain. London, U.K.: World include cryptocurrencies, data science in STEM education,
Scientific, pp. 1–20, 2022. Web 2.0 technologies, ethnoscience, physics teaching, artificial
14. I. Basieva, A. Khrennikov, and M. Ozawa, “Quantum-like intelligence and computer algebra in physics, and quantum
modeling in biology with open quantum systems and field theory. He received the Ph.D. degree in theoretical physics
instruments,” BioSystems, vol. 201, 2021, Art. no. 104328. from the Brazilian Center for Research in Physics. Contact him
15. B. Swingle, G. Bentsen, M. Schleier-Smith, and
at renatopsantos@ulbra.edu.br.
P. Hayden, “Measuring the scrambling of quantum
information,” Phys. Rev. A., vol. 94, 2016, Art. no. 040302.