ad-free
Channel / Source:
TEDx Talks
Published: 2012-06-22
Source: https://www.youtube.com/watch?v=cugu4iW4W54
this gets more in school room foster in foster have you ever wondered when is it ever going to win or one person it's been looking at the miniaturization of computers Avalos several decades has been Gordon Moore and he's the co founder of Intel back in the nineteen sixty and he noticed that the number of components on a silicon chip double roughly every eighteen months to two
years now for this to happen it means that the sports features size on a silicon chip has to decrease the same right he came up with something called Moore's Lauren heritage represents on the screen but this law has been going now for parts before to five decades it was solid out as an observation by Gordon Moore has now become a door off to his name morsel
it's actually continued in time the interesting thing is that the industry is now so this is the road map of how to make computers Morris form foster a foster so you have multi trillion dollar industries the semiconductor industry's pouring money in every year to try and beat that little and so now it's become a self fulfilling prophecy so if we have a look where we are
at the moment here user across sectional scanning electron microscope image the single transistor Abbas Morris feature size in this transistors the distances between the source image right about thirty nanometers it's five thousand times smaller than the width of a human head what's amazing about that as you look around you now we all carry around a personal electronics and within one silicon chip you have a three
billion of these transistors and they will have to work reliably so that your computer your mobile phone whatever you've got with you actually works it's quite amazing just think about that now everybody in this audience has got billions of transistors there are trillions of transistors in this room one of the nice things about Moore's law is you can actually predict with time what's going to happen
and eventually you'll see out here in roughly twenty twenty less than ten years away from where we are now the size of a transistor we get down to size Ricci singleton Festus once compiler of nature it's very difficult to imagine that you could make a transistor any small of let but this is the world of it's true information slips understand how about transistorized here we have
a silicon substrate us with trance that might've been above that we have an insulating oxide and then a metal gate what we do is we apply a positive voltage this top gate here and that sucks up attracts all the electrons they're in Silicon up towards his Cape I can't get a huge this insulating oxide so I formed this two dimensional sheet its forms conducting channel between
source and drain in that sense the transistor on it is all one of ditch to information if we now put a negative voltage in the skate will repel the electrons down here and he pushed me away from that channels there is no conducting cheek and as a consequence you get zero of digital information plus the ones and zeros as we go down for everything that works
around is not everything is coated in either one or zero but what happens is we get smaller and smaller and in science if we actually crossover for we call the classical age to the quantum marriage okay things really start to change in the classical well we understand how things work if I had attended school now announced a threaded war it would hit the wall and it
bounced back and I understand I see it and be able to write quite his emotions to describe that because I miniaturize things down and imagine that tennis for being the electron in my transistor if I made it very very small and I threw that election of the war if if it bounced back it actually behaves walk away from a possible and it can tunnel through the
wall it can come out the other side but something that's quite scary heavy makeup devices will once more out the Wonderful World of quantum mechanics comes in electrons behave like waves and I know that goes in the computer where we want them to guy so a lot of people have predicted that this would Herald the end of Moore's law but in reality is the start of
something you are not transitioning to quantum mechanics and every control quantum physics we could actually build computers in the quantum regime that are predicted to have exponential speed up either classical computers also one of the questions a lot of people lost maison computers foss enough already come they do all the things that we need them to do once everyone wants things before school time but there
are some problems out there that just cannot be solved efficiently using a classical computer I'm one of those is something called the traveling salesman problem so here we have a salesman who wanted to travel to Lhasa different cities and we want to account with the shortest possible riches that sounds like an easy problem protecting one of those intractable exponentially heart problems so here we have on
the screen the number of possible routes they can type has a function of the number of cities it's something that grows very very quickly so for the time you have fourteen different cities there are now already ten to the power of eleven possible routes that you can type so I take a class for computer it works in the gigahertz for jeans pants then on operations per
second and it can work out the shortest possible written about a hundred seconds but it's no big deal but no happens when I catch twenty two cities there are now ten to the nineteen possible routes that salesman can tie and with that same class with computer to take sixteen hundred years this is amazing and if you don't buy twenty eight cities is longer than the lifetime
of the universe to work out what the shortest post reduce service I heard this from many years ago and I just couldn't quite believe it this is a real problem exists out there Sir how can we make a computer that can somehow sold those kind of problems but we have to look at how class with computer work across computers very false but it searches through the
possibilities one after the other rather like a recipe survivor somewhat down a telephone number in a piece of paper not forgotten his telephone number dates or get my class with computers a stock looking through the eyes and then will the peace that overseas eventually would find his number doesn't tell me if I wanted to go to foster I put two computers on to the problem get
one searching three nights wealthy up between upset and that guy foster Dr fost I have three computers what's the ditch to wealth if you could make a quantum computer the actual calculations are done in parallel there will done simultaneously that's why I understand this I'm gonna go back and describe what a classical computer looks like in my mind some medicine I'm sitting at the center of
the earth I'm pointing towards the North Pole we would have talked about North Pole this morning that's my one of pitch two information I could also be pointing at the South Pole best zero of ditch to information but in the quantum world I could be pointing to anywhere on the surface of the if I can be pointing to London by can be pointing to take care
now as a consequence I mean was quoted Pacific position partly up and partly down and that's intellectual might function that's a quantum welcomed in both states the same time but how does that help me and calculations let's look what happens is I crease the number of quantum bits of cubits yes a transistor score in the quantum merging so we just wanna keep it I mean two
possible states the same time if I know what another quantum bit I can be in four possible states the same time ever have another Clinton bit I can be a possible sex the same times every time I add a quantum bit to a quantum computer I doubled the computational power so it's been predicted by having just a thirty cubic computer will be more powerful than was
most powerful supercomputer that exists if I could have three hundred cubits they'll be more powerful than all computers in the well connected together I just stand by for second three hundred quantum bits of Cuba's compared to three billion conventional transistors that's really the power of quantum computation so let's consider some of the problems that quantum computers console for us one of the first things that people
realizes it could actually be useful for data encryption but I transcription relies on working out what the prime factors of a large number on there we have to prime factors and to remind your prime factors in number that can only be divisible by itself or one if I times these two numbers together is a very easy problem for computer can work out your calculator in seconds
but I want to work it with the prime factors of a large number of sixteen very difficult problem rather like the one I've just strategy this underlies the difficulty of problem underlies statesman Christian so we do swing coat our information a very large number and we give somebody one of the prime factors is a key so they can decode the information the other side if they
don't have the key though they have to account for the prime factors are that's very difficult to give you an example very recently they've broken the code in two thousand ten of our save seven hundred sixty eight as a seven hundred sixty eight bit number and it took them three years using the most powerful class computers that existed now encoding is a thousand twenty four bit
number and using the same class with computers it will take three thousand years if you had a quantum computer because solved in minutes so there's an example of how Clinton Paetus when they realize I gotta change the way that we do computing let's look at some other examples I've talked about data security another thing the quantum computers a greater is searching large amounts of data bases
large amounts of information well for modeling system for those lots of variables so you can start to imagine climate modeling mowing the economic system because thought imagine how chemicals form how reactions form how new things start to evolve how human body forms where quantum computation ficus is something we just that night but it has huge potential other consequences a massive international race to build a quantum
computer I don't buy privacy that here in Australia we've decided to do this in Silicon the reason why we've chosen silicon and silicon is one of those great materials the industry's been using it for years if we want to make a quantum computer and silicon graphics engine of single atoms but not just a single atom but the electron the individual action on a single atom in
silicon and encode our information that quantum bit of Cuba a silicon is great because the industry's been working on a fetus but it means that we can be pushing the end of Moore's lost micros singleton transistors so it was also great because it's a material it doesn't interact with electrons it's a nice pure host material to protect that federal quantum state but to realize this one
computer we have to put these individual atoms imposition within a silicon crystal and then we have to align electrodes to the US indexing awesome which means everything has to be incredibly small well how do we image Omineca atoms many technology that exists out there is a scanning tunneling microscope this is something that has a very fine metal tip that it brings down to atom surface when
you bring it down very very close you apply a voltage you got a current were you trying to keep up current constant I move the tip of the atom and as it moves it deflects in high a form that you can actually image the atoms on a surface many restless gannet rather like a television screen and you can build up an image what the atoms look
like on the surface no I'm blown away by this image these individual actions here of silicon sitting on the silicon separates the transistors and might on it's really phenomenal well you might imagine machinery that you use tracks image those items is very small is the very small to but in reality this is what it looks like the very large systems that take up outside of a
room the best be huge chunks of stainless steel with a very high vacuum inside run a lot of actually finding out of space and within that vacuum you put yourselves any controlled the atoms and have mechanical control pumping controlled electronic control to be empty image those items on the surface you can also connect them here with crystal growth systems request put different from some of the
surface you can actually start to create new materials but just don't exist in nature it's given on terror of one of those his the welts more slugger visa xenon atoms on a copper surface is done in the nineteen nineties by the IBM group but on Nightline usually I use that to pick up individual atoms and put them down to form the welts more sloka what we
want to do now is to make devices in silicon using this technology but it's not as easy as just manipulating that happens on the surface it's to keep problems the first one is you can only see the device inside these new systems in Sunday's microscopes you don't see them much you take him outside so we've had to develop that technology the second one is the ecologist
Monica atoms in silicon very easily they actually bone very strongly together so the consequence we have to come up with a radical strategy to build these devices the first thing we do is we have to make a marker in the silicon substrate before we put it into the system we then take it in that we put down a layer of hive Joan surface I miss Larry
Hodges next is a mosque we gonna come along with us getting prepped to absorb some hydrogen bit by exposing the silicon underneath this is how we're gonna bring our items in we dice with fostering that brings office for Saddam's in these are gonna be up cubits I'm Mary Karen these regions that we've anti pacified we then incorporate them into the surface we encapsulate my silicon sonata
bus we gotta go back and imagine that this point show the have moved then once we've done this we take it out the banking system we use as registration markets to bring down metal contacts the device so even when we first presented this proposal back ten years ago little people serve look none of that's been realized each one of those stages is incredibly hard to do
but increases sales trainer of the last ten years we've actually started making these devices systematically building bit by bit those components of a quantum computer we fight might very narrow conducting wise one atom tool room for some wine every Swiss the couple wise you having conventional transistors we've made this most precision transistors would be out to watch individual lex wants hop on and off and on
and forced herself to we've been able to move to three dimensional architectures and for me this is amazing or the transistors we have at the moment or the electrons travel in one two dimensional plane we don't use that direction at all so we found a way we can actually make Vesco transistors something we can use for a computer architecture wanting computer architecture and also very recently
we've been able to isolate a few of these for what happens we've actually used another transistor patented by directly measure the electron spins so we can read out the quantum sites not Clinton Peter but perhaps one of the most difficult challenges frost today has been to isolate a single atom in a device and just last year we were able to form the world's first precision singleton
transistor so it really is an individual faucets Adam sitting in the silicon substrate we've aligned these electrodes to it taken out the vacuum system might contact to it we can actually measure of electronic signature of a single atom directly so all the like the human beings in this audience reach have well defined I