KAY’SCORNER
10 JULY2020|COMPUTER SHOPPER|ISSUE
Tech together to beatthe virus
WHENIFIRSTencountereda
coomputer,it occupied alarge
rooomwithitsownspecialair
coonditioning,andateamof
opperatorstokeepitrunning.I’m
nootsurewhichIBM 360 modelit
waas, but the Model30version
haadamaximumof64KBof
memory,andcouldperformupto
344 ,5 00 instructionspersecond.
Thhatwasmainmemory–RAM.
Thisisn’ta‘backinmyday
hhendinosaursroamedthe
woorld’piece,however.WhatI
wantyoutothinkaboutisthe
computingpowerwenow have by
comparison, even on our phones.
Atypicalmodernsmartphone
probablyhas 8GBofRAM, or to
keep the same units as used so
far, 8,000,000KB, around 125,
times more than thatmainframe
computer. Asmartphone can
run around 4billioninstructions
asecond, 115,000 times more
than the mainframe.
I’ve been thinking about this
because of awhole setofnews
about crowd-sourced or citizen
science. The idea of crowd-
sourced computing is that all that
computingpower in our pockets
or ourPCs can be combined to
create virtual supercomputers.
Most of us have vaguely heardof
asupercomputer, probably a
Cray, costing over £100m to buy
and around £5m ayear to run.
Whatresearchers workedout a
while ago was that if they could
use thesparepower of personal
computers and mobile phones,
they could create virtual
supercomputers that would
give them powerequivalentto
arealsupercomputer.
Perhapsthe best known
example of this is SETI@home,
whereSETI stands forSearch for
Extra-terrestrial Intelligence.This
project from the Universityof
Berkeley,California,hasbeen
lookingforlifeonotherplanets
since1 999 .Over5millionpeople
acrosstheworldhavetaken
partbyinstallingsoftwareon
theircomputer thatruns in the
background. Spareprocessor
time is used to analysedatafrom
space telescopes forsignalsthat
might be radio messages from
other lifeforms in the universe.
ANTI-VIRUSPACKAGES
SETI@home developed into
BOINC (BerkeleyOpen
Infrastructure forNetwork
Computing), which uses the same
techniquefor awide range of
distributedcomputing projects.
While the diseases being studied
by BOINC projects have until
recently beenmajorkillerssuch
as malaria, HIV andcancer,the
emergence of Covid-19 has driven
researchers on its Rosetta@Home
project to work instead on
predicting the structure of
proteins important to Covid-19,
and to look fornew mini-proteins
that mightbeusefulin
diagnosing and treating thevirus.
This is just one of anumber
of studies you can joinwith
either your personal computer
or your mobile phonetohelp
discoveravaccinemore rapidly.
Folding@Home is another
long-running project thathas
also moved to trying to finda
cure forCovid-19.Folding@Home
is nowlookingathow viral
proteins work and how to stop
them working, which wouldbe
away to prevent Covid-
suppressing the immune systems
of people infectedwith it. The
researchers at Washington
University,StLouis,say their
community-based distributed
network is nowrunning at over
1,000petaflops, over five times
quicker thanthe world’s fastest
supercomputer, andthat makes
it the first exaflop computer
(an exaflop is 1,000 petaflops).
Another call foryourunused
computerpower is slightly
different, in that it’saimed at
yoursmartphone.Vodafonehas
createdanappcalledDreamLab
forcancerresearch.Thecurrent
situationhaschangedthefocus,
andthenewlylaunched
Corona-AIproject,acollaboration
betweentheVodafoneFoundation
andscientistsatImperialCollege
London,istofindtreatmentsfor
coronavirus.Itisn’tlimitedto
Vodafonecustomers;anyone
with asmartphonecan join. You
download theDreamLab app,
then set it to run at night while
your phone isn’t being used. This
uses data, but youcan set it to
run on aWi-Fi connection, so
avoiding usingupyour monthly
allowance.The app uses around
500MB of dataamonth,
assuming it runsfor sixhours a
night, and no personal data is
accessed or affectedinany way.
Theapp has beendownloaded
more than 100,000 times, and
the team from ImperialCollege
saythat if that many people run
it forsix hours anight forthree
months, it will enable them to do
in that time what it would take
the supercomputers at Imperial
College ayear to process.
DATA CRUNCHING
Corona-AI is working its way
through avast amount of data
to see if there areexisting drugs
that look as though they could
be effective againstCovid-19.
Existingdrugs areofparticular
interest to the team because
they’ve already gone through
the testing stage.Dataon
food molecules is alsobeing
searched,tosee if there are
foods that might help because
of anti-viral properties.
I’ve beentempted in the past
by these research projects, butit
seemed afaff,and I’mnot that
bothered about contacting alien
lifeforms. Coronavirusisrather
more urgent, and if we can
help by letting researchers
use ourunused computer or
smartphone power,itseems
asmall inconveniencefor a
potentially massivereturn.
KAYEWBANK
SoftwareguruandShopperlegend
[email protected]
Crowd-sourcedcomputing projectsareharnessing thepower of PCsand smartphones
inthebattleagainst Covid-19.KayEwbankexplainshowyou canhelp
Thereare anumber of studies youcan join
to help discoveravaccine morerapidly
