80 Science & technology The Economist December 4th 2021
Another member of the project, Shash
ank Misra, says coinflips’ researchers
have identified two hardwarebased ap
proaches for the production of tuneable,
abundant random numbers. One relies on
the patterns magnetic films make when
disturbed, the other on how electrons tra
vel through the barrier of a quantumtun
nelling diode. Both of these things are truly
random. And both can be tuned to provide
the sort of randomnumber distributions
coinflipsrequires.
In a quantumtunnelling diode, elec
trons randomly leap (or fail to leap) across
a gap, and the distribution of success or
failure can be changed (and therefore
tuned) by altering the voltage. A magnetic
film, meanwhile, is composed of many ti
ny magnets jostling with each other. Flip
the polarity of one of these and others
around it flip in response, creating a ran
dom pattern. In this case the plan is to
build tuneability into the magnetic medi
um itself, by tinkering with its composi
tion, and also applying strain to it.
Randomnumbergenerating hardware
based on either of these approaches could
be built directly into chips. Alternatively, a
randomnumbergenerating unit might be
added to existing devices to yield an end
less supply of highspeed randomness in
the way that graphicsprocessing units are
added to machines that have to handle a lot
of graphical material. Crucially, unlike the
heatflux method, which requires inter
mediate hardware involving thousands of
transistors to translate the signal into digi
tal bits, the results of the diode and film
methods can be read off as bits directly, us
ing only one or two transistors to do so.
The efficiency offered by coinflips
means some tasks which currently require
a supercomputer might be carried out with
desktop hardware, depending on how
much of the load involves generating and
manipulating random numbers. Dr Ai
mone says the current approach tends to
be simply to build bigger computers where
needed. But even then some largescale
tasks may be too costly in money and time
lost to conduct. You can, for example, run a
model of a hurricane’s path only so many
times before the real thing makes landfall.
The heart of the matter
Despite its peregrinations elsewhere,
though, Sandia is ultimately in the nuclear
business, and one early application of
whatever coinflipscomes up with is like
ly to involve interpreting the results of col
lisions in particle accelerators—some
thing the team have been exploring in col
laboration with Temple University, in Phil
adelphia. The idea is to build a device
which incorporates coinflips hardware
into the sensor itself. This will allow re
sults from collisions, which will be ran
domly distributed, but in particular ways,
tobecomparedwithartificialrandomdis
tributions,toseeiftheymatch.Tobeable
todothisinrealtimeisuseful,becauseit
allowsanimmediatedecisiontobemade
aboutwhetherornottostorea particular
result.Moderncollidersgeneratesomany
collisionsthat suchimmediacyin deci
sionmakingisimportant.
Thatisone,ratherspecific,application.
But inthelonger run, says Dr Aimone,
coinflips shouldenable many typesof
calculationthatarecurrentlyimpossible
becauseofthevolumeofrandomnumbers
needed—for example, artificialintelli
gencesystemsthatcapturetheuncertainty
oftheworld.Thismightbedoneinthe
formofneuralnetworks which,likethe
humanbrain,haverandomnessavailable
ateachsynapse.Andthat,inturn,maylead
tocoinflips, a projectinspiredbybiology,
returningthecomplimentbyprovidinga
betterwayofunderstandingofhowbrains
themselveswork.n
Planetology
The memory
of water
E
arth—the quintessential blue planet—
has not always been covered by water.
Around 4.6bn years ago, in the solar sys
tem’s early years, the energetic young sun’s
radiation meant the zone immediately sur
rounding it was hot and dry. Earth, then co
alescing from dust and gas in this region,
thus began as a desiccated rock. How it
subsequently acquired its oceans has long
puzzled planetary scientists.
One possible source of Earth’s water is
carbonaceous (ctype) asteroids, the most
common variety. But it cannot be the sole
source, because water in chunks of these
that have landed as meteorites does not
match the isotopic fingerprint of terrestri
al water. This fingerprint is the ratio of nor
mal water (H 2 O, made from hydrogen and
oxygen) to heavy water (D 2 O and HDO,
which both include deuterium, an isotope
of hydrogen that has a neutron in its nucle
us alongside the proton characteristic of
every hydrogen atom). Water from ctype
asteroids has more deuterium in it than
does terrestrial water.
Another possibility is comets, which
are basically dirty snowballs that arrive
from the outer solar system. A barrage of
these a few hundred million years after
Earth’s formation would have done the job
nicely. But samples returned from comets
by spacecraft suggest their isotopic finger
print is even less Earthlike than that of c
type asteroids. So, as Luke Daly, a planetary
geoscientist at the University of Glasgow,
in Britain, observes: “It basically means we
need something else in our solar system,
some other reservoir of water to be on the
lighter side to balance the books.”
In their search for this reservoir, Dr Da
ly’s team recently studied grains of silicate
dust from another samplereturn mission,
to an asteroid called Itokawa(pictured).
This is an stype (stony) body, with a com
position different from that of ctypes. The
grains had been brought back by Hayabusa,
a Japanese craft, in 2011.
Grains of this sort formed at the same
time as Earth, and then spent the interven
ing billions of years orbiting the sun, occa
sionally gathering into tiny rocks or falling
onto the surfaces of asteroids, such as Ito-
kawa, to create a finegrained regolith.
Most, though, have remained freefloat
ing. Indeed, they are collectively visible at
sunrise and sunset, in clear, dark skies, as a
faint glow known as the Zodiacal light.
Using a technique called atomprobe
tomography, Dr Daly was able to examine
the composition of the grains in his pos
session one atom at a time. He found, as he
describes in this week’s Nature Astronomy,
that they contained a significant amount
of water just below their surfaces. That sur
prised him. What was intriguing, though,
was his discovery’s lack of deuterium.
Dr Daly reckons this water’s existence
can be explained by weathering of the
space dust over billions of years by the so
lar wind, a stream of charged particles—
mostly protons—that flows out into space
from the sun. When they hit a particle of
space dust, these protons penetrate a few
nanometres below the surface and change
its chemical composition. In particular, if a
proton knocks out one of the metal atoms
in a silicate’s crystal lattice, it is then likely
to bond with an adjacent oxygen atom to
form a hydroxide ion (OH). Add a second
To find the origin of the oceans, look
in outer space
Acosmic reservoir