120 AUGUST 2019|COMPUTER SHOPPER|ISSUE 378
QUANTUM CRYPTOGRAPHY
One of the best-known algorithms that could
take advantage of aquantum computer is
Shor’s algorithm. While factoring anumber
using aconventional computer is essentially a
trial and error approach, Shor’s algorithm
would make light work of it.
This might be acuriosity to
mathematicians, but the impact could be far
reaching. The most commonly used forms of
encryption rely on the fact that factoring a
large number would keep the fastest computers busy fordecades.
But if practical general-purpose quantum computing were areality,
cracking encoded messages would be much more achievable.
Quantum technology might render today’s encryption
technologies obsolete, therefore,but it could also offer an
alternative,which is mathematically proven to be 100% effective.
Enter quantum encryption.
Strictly speaking, this isn’t aform of
cryptography so much as amethod of
distributing acryptographic key. Any
encrypted message is totally secure if the key
is the same length as the message and is
never reused, but this has generally been
considered impractical because of the
difficulty in one party securely transferring
the keytoanother.However,the quantum
approach allows two parties to share akey
using astrange quantum effect that tips off the communicating
parties if an eavesdropper has intercepted the key.
Unlike the general-purpose quantum computer,which is probably
at least adecade away, quantum cryptography is with us today.
While any government and military systems are likely to be
shrouded in mystery,it’s interesting to notethat several companies
already offer commercial quantum keydistribution systems.
at least one of the coins wobbled and
fell to the table.Even if we somehow
managed to get 300 coins spinning on a
tabletop at the same time,the slightest
bump of the table would cause all of the
coins to collapse.This is the challenge
of radio frequency noise in aquantum
system. To eliminatenoise,qubits
must operateatatemperature of
approximately 20 millikelvin, 250 times
colder than deep space.”
All of this begs the question of
how close we are to apractical
quantum computer.Clark’s analysis
isn’t tooencouraging.
“When it comes to developing a
commercially relevant quantum
computer,we’re at mile one of a
marathon,”hesuggested, before
referring to apossible target.
“Commercially viable systems will
require tens of thousands or close to
one million entangled qubits.”
WAYTOGO
Meanwhile,Filipp’s views on how many
qubits are needed in areal-world
quantum computer,compared to
today’s 50, seem to support Clark’s view
that there’s along waytogo, albeit not
quiteasfar as Intel’s prediction.
“Wehavealready implemented a
new quantum algorithm capable of
efficiently computing the lowest energy
stateofsmall molecules,”Filipp says.
“More recently,our team in Zurich
has published aquantum algorithm
that analyses risk more efficiently
than MonteCarlo simulations
traditionally used on classical
computers, which has implications for
financial risk analysis. To achieve more
complex applications, we will need to
reach afew hundreds of qubits.”
We might only be amile intoa
marathon, and today’s 50 qubits might
look rather puny compared to the
hundreds, hundreds of thousands, or
million qubits that we might need, but
what does that mean in terms of time?
IBM believes we’ll get there in the next
fewyears, although Intel pictures us
currently being where we were with
classical computers back in the 1970s.
“In almost every way, the outlook for
quantum mirrors the development of
supercomputers40yearsago,” Clarksays.
“The introduction of the Cray1in
1975 marked the most advanced
commercial computational system
ever available.Every national lab in
the country bid to be the first to
receive the machine.
“But today, the smartphone in your
back pocket has more computing
power than researchers could have
ever dreamed of fitting intothat
5.5-tondevice.Today, the outlook for
quantum is big, bulky, cold and complex.
But in the near future,theycould
provide exponential computing power
to every technology-driven task we can
dream up,first in research, like in
national labs, then in business.”
But who’s going to benefit? Are
there really multitudinous applications
forquantum computers? After all, what
we mostly seem to hear about is Shor’s
algorithm forfactoring large numbers.
Filipp begs to differ.
“An increasing list of applications
can be found atquantumalgorithmzoo.
org,and these can be applied to
problems ranging from quantum
chemistry applications to tasks in
machine learning,”hesuggests.
“Weare exploring how these can be
applied to real-lifesituations. We have
clients in the automotive industry,
including Daimler and Honda, and in
finance such as JPMC and Barclays.”
Given such abroad range of
applications, it seems barely
conceivable that the quantum
computer will be atechnology looking
foranapplication. Indeed, we can
imagine the next generation questioning
how we ever lived without it.
ABOVE:So far
there’s been no
consensus on the
best waytobuild a
quantum-computing
chip.IBM’s
approach involves
superconducting
circuits
Im
ag
e:
IBM
Re
se
arc
h
