Lunch & Learn: Quantum Computing

psychos enabling the digital economy so my program for the day is essentially three things I would like to first of all open the box unpack the quantum computer I'm sure you all get exposed to media press articles and videos about how weird and spooky everything quantum is and nobody understand it it's all mysterious that's not true okay I don't have the ambition of teaching your quantum mechanics in 42 minutes but I would like to demystify DDD you know the complexity of it and being comprehensibility of it it's really not that odd and then I will go into the actual meat of the of the presentation which is to give you a sense of what a quantum computer is how we build one and a few example of what it does and what it can be useful and I will touch upon at the end and also throughout the talk my own personal activity here at the University of New South Wales in Sydney on building quantum computers out of silicon now before we start let's really begin from the from the core of the matter what is quantum what does that even mean right and the correct answer is that quantum is the description of the natural world that accounts for a essentially a law of nature a new law of nature like Maxwell's equations like Newton's equations are laws of nature and it's called the Heisenberg uncertainty principle and it says what you see there on the screen that the product of the uncertainty on the position and on the momentum which is the velocity times the mass of a particle must be larger than half the Planck constant which is a constant of nature that's about 10 to the minus 34 joule-seconds now I don't expect this to mean anything to you so let me give you a practical example all right so let's take a tennis ball let's say you're a tennis player and I'm taking an amateur tennis player hits the ball at about 100 kilo there's an hour so let's say 30 meters per second the ball is about 50 grams so for a tennis ball the Heisenberg uncertainty principle says that if you want to know its velocity to within 10 to the minus 16 meter per second so 17 significant digits of precision z' of their velocity the uncertainty on where the ball actually is has to be 10 to the minus 17 meters which means less than the size of an atom in fact less than the size of a proton so this is the sense in which we say a tennis ball is a classical object is not a quantum object it's actually a quantum object - but because of its size and mass the quantum uncertainty principle doesn't matter it's completely irrelevant let's take another example let's take an electron at the same speed the electron is much much lighter than a tennis ball of course so if you want to know its velocity to within 30% so a 10 meter per second uncertainty on the velocity you need to have a 10 micron uncertainty on its position which is much more than its own size which means basic you have no idea where it is so in this sense an electron is a quantum particle is the same law for the tennis ball in the electron but for the tennis ball it doesn't matter it's irrelevant that uncertainty is so small that it doesn't matter for the electron it completely rules its world so because of this uncertainty you get funny effect which are completely natural such as superposition effect so I imagine you have an electron gun you shooting electrons like you will like the gun that shoots tennis balls when you're training right but it does it with electrons and then you have a screen with two slits and those two slits are less than 10 microns apart okay where does the electron go the left of the right slit what you don't even know where it is - better than the distance between the slits so the only logical answer is that the electron goes through both slits that's all you can possibly say and then when you look behind the screen there's going to be a interference pattern that looks like the pattern of a wave so that's what you would have you tried in your own bathtub to make a cardboard with two slits and drop a stone you'll get two waves that go towards the slit and you'll see an interference pattern behind the cardboard double slit now you can actually see these electron waves with certain tools that are themselves quantum tools is a scanning tunneling microscope experiment of the surface of a piece of copper which is a metal that has free electrons everywhere when they've made a Corral of iron atoms that Corral confines the electron waves inside it and so you can actually see this ripple this wave of probability of the electron being confined by this Corral is actually a fantastic you know modern technique to see the quantum world and it's wave nature now because of superposition has a lot to do with symmetry so let's say you have two protons two positive charges that are identical and you have one electron okay so the electron is a negative charge would like to bond to the proton to the positive charge now which one does it choose does it choose the left one or does it choose the right one what do you think is the logical answer both why not we choose one of them the logical answer is that it would be in a superposition of being left and right there is nothing weird or odd about this and it seems like you all agree with me I'm very happy and so that the images here at the bottom is a story presentation of the electron probability distribution around a pair of protons in a hydrogen molecule for example let's look at another example of a quantum particle as the spin every fundamental particle has a spin so protons neutrons electrons ever spin spin is essentially a microscopic magnetic dipole like the needle of a compass but really really small at the microscopic scale and so it produces a small magnetic field and you may wonder can we see a spin can we actually observe the magnetic field produced by the spin and if you've ever had the misfortune of needing an MRI scan in horse pittle you would have seen the concentration of proton spins inside your body so this is an MRI scan and what you see the light contrast means that there is a high density of proton spins in there so you can actually see it and it's a diagnostic medical tool these days but that's a lot of spins can we see a single spin this is hard but it can be done and I'm gonna show an example from my own laboratory here so this is a device that was made by one of my students and the colors are fake that just for presentation purposes but this is basically a silicon transistor it does may not look like one but it's a modified silicon transistor the yellow strip is a microwave antenna kind of like the super high frequency and nanometer scale version of this antenna here and then next to that transistor which is colored in red we implant a single atom of phosphorus phosphorus is an automat when placed in silicon captures an extra electron which itself as a spin and also the nucleus itself of the phosphorus has a spin so you get a two spin system in there and now this is a video it's an animation that's been made by a student of the College of Fine Arts but it's actually completely accurate so this is what you would see if you imagine being like a silicon atom inside inside the silicon crystal looking up at the ceiling at the top of the of the silicon shape so the blue you see there is the electrons that have been pulled up by the electrodes of the transistor and then in red there is the nucleus of the of the phosphorous atom and then the orange lines represent an electric field that controls a charge so we start from the phosphorus atom ionize it doesn't have an electron attached to it and then we grab an electron from that pool of electrons in the transistor in the spin down state in the low energy state and then we radiate it if we hit the right frequency the electron will turn up to the high energy State and then from there it has enough energy to escape so it escapes back into the transistor and when it does that it unlocks a burst of current which lasts until another electron in the low energy State goes back onto the atom now we try and irradiate with a different frequency which is the one it would respond to if the nucleus would flip the other way in this animation we haven't flipped the nucleus so the electron does not respond so it does not go to the high energy state so it does not escape and so it doesn't give your current okay so this is a lot of blahblah but let me show you the real thing this is the real thing this is actually the oscilloscope screen in my lab this is a video of taken with my phone okay this is what you would see if you came to my laboratory what you see there is the current through this little transistor you see there our disappearance but there's basically no current is a bit of noise but otherwise it's flat and now you'll see in a moment you get this burst of current this bursts of curve every time you have successfully excited the electron spin and you see you got a lot of bit and then they stop why do they stop because we flick the nucleus the other way so now the electrons no longer respond so now I would like you to ponder on this for a moment okay what you are seeing here you're watching with your own eyes the quantum state of a single electron in real time there's no like this is the real thing then sped it up or slowly down this is what you see in the lab and also you're seeing the quantum state of a single nucleus because you can tell whether there's zero current or a lot of bursts depends on the direction of the nucleus pain okay this is the kind of thing that my quantum mechanics teacher 20 years ago when I was at university taught me cannot be done people really believed that if you lit read you know old papers by you know Einstein and Schrodinger and so on they say oh this is nonsense this can never be done we do it every day it's not because quantum theory itself has changed quantum theory is the same for last 80 years but is the quantum technology that has changed the tools we have to make contact with the quantum world have changed dramatically in the last twenty years this is what sometimes it's called the second quantum revolution the fear is the same but the technology is so much further you can now watch a single quantum object with your eyes every day okay now one more thing about spins well about quantum theory is entanglement this is the thing that broke Einstein mind right here it was the famous spooky action at a distance and you know God doesn't play the eyes and things like this we now understand it perfectly well and we've done experiment to show that this is true and again it's actually very simple when you accept the symmetry of the world so if you take two spins right they each produce a magnetic field on each other and so one makes a feel this way and the other one will make a feel that way so they want to point anti parallel so they could be this way if you put them side by side or they could be this way which one do they choose none of them they are in a superposition of this way and that way and what that actually means is that each one of the spins doesn't actually point anywhere so an entangled quantum state is a state where each spin is opposite to the other but doesn't point anywhere it sounds really strange when you say it in you know normal day-to-day words but it's actually completely normal in fact you are full of this stuff every covalent chemical bond involves a pair of entangled spins that's a water molecule that's seventy percent of what you are in every chemical bond between the oxygen and the hydrogen there is a pair of spins that have that entangled quantum state there is nothing we are the body we are all made of this alright now let's get to the to the actual business side of things which is computers ok that's what you all want to know about so again I'm gonna take it from the from from the very beginning that's an old image of an 82 to 86 from you know the 1980s for those of you remember that so let's say you want to do a little calculation you want to calculate how much is three times five computers that digital computers right so you let's say use four bits so you say three will be 0 1 1 & 5 would be 0 1 0 1 and the computer will calculate that 3 times 5 is 15 which is 1 1 1 1 in digital code now as the great Roland our one of the fathers of information theory both classical and quantum famously said information is physical so if I say I have a zero in my computer what does that actually mean that zero is a physical reality and the physical reality in all the computers you have in your pocket or on your desk is the voltage across a silicon transistor ok so zero means low voltage across the terminals 1 means high voltage across the terminals it's a physical thing and by the way the physical object so the silicon transistor is in my opinion the highest achievement of humankind it's much more impressive than going to the moon and back is much more impressing the finding a Higgs boson if you actually go and look at what's in your pocket there is a billion objects man-made objects on a scale on a size scale of a few hundreds of atoms across from end to end it's a billion of them and they all work on average there's a hundred broken transistors out of a billion transistors chip and you buy them in the shop for for a few dollars this is the most amazing achievement of humankind ok so I'm going to tell you about quantum computers now but keep in mind that what we have in our pockets and on our desk is a fantastic feat of technology but we have other choices so let's look at what happens if we use instead of a silicon transistor a quantum object that has two natural States to encode quantum information so for example I could take two protons and one electron and I say ok if the electron is on the Left I call it a zero the electron is on the right I call it a 1 but as we've discussed before it can also be in fact it would like to be usually 0 and 1 at the same time that's a completely legitimate code on this quantum bit another example is the spin so I can called spin down a zero spin upper one and again another natural state is the superposition of up and down now this in itself doesn't give you very much you only start to understand the power of quantum computing when you look at multiple bits and I'm going to give you an example here so let's say you have three classical bits okay this is a classical case with three bits you can write eight numbers 0 0 0 0 0 1 all the way down to 1 1 1 ok 8 numbers 3 bits in terms of information yes you have a choices but all the information in there is 3 bits of information it to give you the first bit the second bit and the third 3 bits now let's say I redo the same thing with quantum bits let's say I use a spin that's the thing I use in my lab so now I always have those eight we call them basis states like the 0 0 0 0 0 1 and so on but because spins are quantum objects I am allowed to make super positions of all these eight states just like I did a superposition of the spin up and down before now I can make a superposition of down down down with down down up with down up down and so on all eight of them so now if I want to describe what is the quantum state of those a of those three spins I need eight numbers I need to give you the diffraction of you know it's a little bit of this a little bit of that a little bit of that so that's eight numbers where 8 is 2 to the power 3 so if I had 4 quantum bits I would need 16 numbers if I'd 5 I would need 32 numbers so in general if you have n quantum bits you need 2 to the power n numbers to describe the full quantum state now why why is it more than the classical case so why do I need 8 here and I only need 3 in the classical world the reason is actually because when you do this kind of super positions most of the states you get are entangled States so you get a lot of those situations where the bits exist only in relation to each other not in their own individuality this isn't you cannot do on a classical computer but it is a perfectly legitimate digital coordinate in a quantum computer it's as if you had a vocabulary that is exponentially richer and every word is completely legitimate except you cannot write it on a classical computer ok now let's take this to the to the bigger scale so that number there is 10 to the 19 so it's 1 with 90 zeros after that that's more or less how many particles there are in the known universe now 10 to the 9th is about the same as 2 to the power 300 so what that means is that if I had 300 quantum bits there are all perfectly fully entangled to describe their quantum state I would need as many numbers and this as there are particle in the universe it's the power of Exponential's ok now let's do something a little bit more concrete and this until you may have heard of in the news like quantum supremacy or quantum advantage they are actually quite simple let's go a little bit smaller let's take 50 cubits 50 quantum bits so you'll need 2 to the power 50 numbers to describe the quantum state so the code encoded in those 50 cubits now 10 to the so 2 to the 50 is about the same as 10 to the 15 so imagine use one byte of classical data for for every coefficient so it means you need a petabyte a petabyte is a thousand terabytes and that's something you can kind of visualize and I imagine your typical you know terabyte hard drive that you have to back up your data imagine stacking a thousand of them here on the stage you know it'll make a little pile that's a petabyte okay you can kind of vision that now let's go to seventy cubits that's so you need to to the 70 which is about 10 to the 21 so that's a zettabyte so that's a bill terrabytes so now imagine a billion of those terabyte hard drives they will overfill this room that will overfill probably the convention center and in fact there is about as much digital data as there is in the world at the moment take everything like all your laptops servers data centers the whole shebang that's about a zettabyte of that everywhere so now you see that if you could make a quantum computer with seventy cubits there are all fully talking to each other and entangled with each other and you could run a calculation on this computer that uses the entire 2 to the power 70 information content of the computer you could legitimately say that you've done something that cannot under any circumstance be done on a classical computer now these are very oversimplified argument but it kind of gets the numbers more or less right so you may have heard you know Google is now they got a 72 qubit prototype that kind of reach in that scale that's where we're at this is the scale where you can legitimately say we're doing something that cannot even in principle be done on a classical computer this is about to happen so stay tuned now what what do you use this for okay what are the applications of quantum computers and I'm gonna give you again because I like doing things simple a simple example of the kind of problems that quantum computers are good at solving they are the ones where the computational complexity explodes exponentially that's what they're good at so here's an example of a combinatorial problem let's say you have two eggs and a six-pack of eggs how many different ways can you put two eggs in a six-packs so for the first tag you've got six choices for the second egg you've gone lega five choices because one you've plugged before and then if the two ways you consider them identical it means you can swap them around and get the same configuration so you divide by two so in totally got 15 different ways to put two eggs in a six pack if you do four eggs in a 12 pack now you've got 12 choices for the first one 11 for the second time for the third nine for the fourth and then you have to divide by the number of permutation of four eggs which is 4 times 3 times 2 so that's 495 choices of how to put 4 eggs in a 12 pack so it goes up pretty quickly and then I have a friend who works in a cafe around here and so I asked her to take a picture of 20 eggs in a 30 pack the stuff they have they're a cafe and so you do the same number calculation there and you get 30 Millions you see how the complexity of this problem absolutely explodes as soon as you start and make it bigger now this is not the kind of problem you want to solve on a quantum computer as such but I give it as a sort of metaphor of the problem of quantum chemistry and molecular simulations so let's say you know you've seen these images of you know atomic orbitals and electrons in there so you can kind of imagine that the electron is like an egg and your beetles of the atom or the molecule are like that the pack of eggs so what you're trying to do when you do molecular chemistry and quantum chemistry is try to figure out what is the configuration of electrons that has the lowest energy so the one that is most stable and so I hope that my little you know cafeteria example gave you a sense of how complex this is there are so many different configuration even just with a few orbitals all right so here is an example that's been done very recently at IBM in fact some very simple molecules called beryllium hydride it's one beryllium atom which is element number four so got four electrons and two hydrogen atoms very simple molecule so a sixth electrons in total now for reasons I won't go into if you actually did the full quantum mechanical calculation of the energies of each one of the possible configurations of the electrons in this molecule you would have to solve a 40,000 by 40,000 matrix this can be done but it's a heavy calculation already so what people have done at IBM is that they've mapped they don't kind of the eggs in a carton example I've I've shown to you so they've mapped each electron onto one of their quantum bits this is a processor with several quantum bits they only use six out of seven in fact so it's each one of the sixth electrons is one of the quantum bits and then they arrange the interactions between these quantum bits in a way that resembles the interactions between the electrons in the beryllium hydride molecule and then they get essentially a a result so the configurations of this quantum bits tells you Maps directly on the most convenient configuration of electrons in the molecule okay so instead of solving a forty thousand by forty thousand matrix you just run a seven qubit quantum computer okay now that's what I use less molecule barium hydride no one cares about it but let's say you go to something like this this is actually again a simple molecule is penicillin it's one of the simplest you know drugs this would be completely out of the question for a classical computer you could never solve this exactly there are in fairness some very good approximation methods that people in quantum chemistry use to get a sense of what goes on here but fundamentally the reason why you cannot just go on a computer and design a drug that does something for you is because that problem is computationally intractable it's just way too hard because it's a quantum problem because it's the problem of the quantum mechanical lowest energy state over small but it doesn't need to be that big to be already intractable number of quantum particles so these one of the hopes we have you know the nearest term applications of quantum computing is for drugs and molecular design and material design and so on so problems that are intrinsically quantum that easily become intractable on a classical computer and then a quantum computer is sort of the natural platform to address them so these are problems will of course very big impacting society and in business let's look at another actually really simple molecule ammonia one nitrogen three hydrogen's so simple and innocuous but ammonia is what's used for fertilizers so we would not have seven billion people on earth if we didn't have a way to make ammonia for fertilizers to feed the world except the production of ammonia consumes about two percent of the world's energy that's an ammonia plant in India it's a very inefficient process it requires very high temperatures high pressure it's slow it burns a lot of energy to make a mone of course we need it so it's energy well spent now you think isn't that a better way and you might think well maybe there isn't a better way maybe that's just what you need to do to make ammonia but we know it's not true there are bacteria that spontaneously produce ammonia without having to heat up to 600 degrees so we know that there is a way but we haven't found it why haven't we found it because it's a quantum process that chemical reaction that goes on inside that Sun of cyanobacteria is a series of quantum evolutions and we just can't make sense of them on a classical computer so we hope that quantum computers will be able to resolve this kind of really impactful and an important problems in in science and in economy alright another example which I don't think it's gonna change the world as such because it doesn't go from exponential complexity to polynomial but is the database search so this is one of the very first quantum algorithms that was discovered by grover in 1996 so if you take an unsorted database so let's say I gave you a telephone book and I told you I have this number can you find whose number this is the numbers are not in order and there is no better classical algorithm just crawling through the whole book until you find a number you find a name in a quantum database you can prepare a superposition of all the entries in the database and you can arrive at the result in the square root of the number of entries so if you had a million n let's say the Sydney telephone book with a million names in it on average you need half a million checks until you find a number on a quantum computer you need a thousand checks it's nice ok and I have a video on youtube if you're interested it's called quantum computing concepts at 3 3 minutes short videos and in one of them explained their quantum search algorithm all right now this is all what you've all been waiting for factoring right you guys work in the finance industry you have you know data that is precious that needs to be encrypted and be kept safe and you've probably heard that a quantum computer can crack classical encryption because it reduces the complexity the computational complexity of finding the prime factors of a composite number from exponential to polynomial ok so it be it turns an intractable computation into a tractable one so this is really what you know most of the you know finance and defense and government agencies are interested in because that's the one that can have a massive impact in theory now let's get to the practice nothing in the world is perfect right so in a computer in fact yes there are the bits you're trying to calculate but there's actually a lot of redundancy because you need to have the ability to detect and correct errors so usually for example when you transmit a bit of information you don't transmit that bit of information you may transmit three copies of it so that if there is an error along the line and you get you know one bit is 1 and the other 2 are 0 you can say well there was probably a 0 it should have been 0 0 0 and then there was a mistake and you flick it back you can do something similar with some caveats in a quantum computer so you main code instead of on one bit on three bits and so the single bit of information we call the the physical qubit and the encoded in many physical qubits bit of information is called a logical qubit and so people have invented some very efficient methods do large-scale quantum computations on two dimensional grids of qubits and so the this kind of code and error correcting code there's a tolerance of about 1% so it can correct errors as long as they occur less than 1% of the time which is very good and so an average you would need but if you go below the threshold you need less redundancy so people normally look at let's say 0.1% error you need 3600 physical qubits per logical qubit ok so now let's look at what it mean in practice to find the prime factors of a number let's say you get a 2,000 bits number and you that that is the product of two you know smaller numbers that are some cryptographic keys you want and you want to crack that cryptographic thing so you'll need twice that in logical qubits or 4,000 let's say you get point 1 percent error rate you need 3600 physical qubits per logical qubit so it's about 20 million physical qubits and then there's something I won't go into you need another two hundred million cubits for other stuff believe me let's say you have a hundred nanosecond clock cycle which is pretty good there would be state of the art that means you need 220 million physical qubits and it will take a day to do this calculation ok so as I said before we are now kind of racing at the 50 70 qubit level so this is to say that quantum decryption is not around the corner it's really not around the corner will take us a while before we get there but at the same time there are motivations why you may want to start being interested in that because of course someone could capture your communication stored them on a hard drive and then 20 years from now crack them so depending on the sensitivity of the data you're handling you may want to start caring about this and many people are starting already which is a good idea I have to say also in defense of the quantum technologies the quantum technology can also be used for good so you can use quantum encryption to protect your data this is that Euless from a company here in Australia called quintessence lab and they you offer a suite of of technologies like quantum key communication quantum encryption so these things actually exist and in fairness they are much more advanced than the quantum computing itself so one thing I'd like you to take home is that quantum safety is actually much more advanced than quantum danger okay so we are on your side okay now what does the quantum computer look like these are three examples the one on the left is actually my chip from my lab it's a small silicon chip on a printed circuit board the one the middle is an IBM v cubed superconducting quantum computer the one on the right is an ion trap quantum computer from Austria and the scale of these objects is about a millimeter for my chip a centimeter for the IBM chip and 10 centimeter for the ion trap vacuum chamber so this seems like innocent but then you say ah but wait a minute these things for example my little chip is mounted into a giant one giant is like 2 by 2 by 3 meters thing that cools it down to near absolute zero all right and also the ion trap I mean there's a trap there but then there's this massive optical table that's as big as this stays as your lasers and lenses and so an objection you often hear is that well I mean how am I going to put that monstrosity in my office it's never going to work of course the point is that you're thinking of a classical computer as being a laptop like this which is not the case of course a classical computer is this it's a data center there's hardly anything you do nowadays that actually takes place in your computer it takes place in some data center and so in that data center if you now put to scale my giant refrigerator it's actually looking like this so what we envisage is that there would be data centers that will have additional you know quantum rooms with machines like this and you will connect to them remotely possibly with a quantum secure connection and then you can get quantum calculations down there and you just you know from your own normal lab now I just want to say one thing because my background is in low-temperature physics reaching these low temperatures is actually probably the only achievement of mankind where we can legitimately say we beaten nature so if you're going to empty interstellar space in outer space the temperature there is 2.72 27 degrees above absolute zero because of the microwave cause big background radiation whereas in my lab we get 0.01 degrees so we've done something that nature just doesn't do and we use it to make a quantum computer work all right so last thing so we have a little bit of time for questions so my activity here is to build quantum computers using silicon I told you before that silicon microchips are in fact the most amazing achievement of humankind so the idea was can we reuse readapt these incredible technologies that trillion-dollar industry making these chips to make quantum computers so what you see on the left is actually a fairly obsolete from 10 years ago 45 nanometer node silicon transistor what you see on the right is a quantum bit two qubit device made by my student a couple of months ago and I mean that that's a cross-section the one is a thing from above but actually these things are quite similar they have basically the same size and they're made with the same technology mean if anything the technology I use is much more primitive I don't have of course the facilities of an Intel factory right I mean I'm working at University Research Lab so with this kind of technology we have in fact achieved the longest quantum memory time of anything in the solid state using this phosphorus atom so this stuff works really well and we are now developing new technologies new schemes to make that remember that two dimensional grid I showed you before the surface code to actually have these atoms on a two-dimensional grid and operate them in order to perform large-scale quantum computation so this is an ongoing project we're doing very well here in Australia and there's other people around the world also racing with us but it's a good healthy race so with this I'm going to finish I just want to re highlight a few key messages that I would like you to take home first of all is that quantum mechanics is just the way things are okay don't let anyone tell you that oh is this weird spooky counterintuitive no it's just the way things are you are a quantum object we all are everything is quantum okay and in fact it's got great potential for novel technologies we can do things with the quantum property of particles that are very useful quantum information is exponentially more dense than classical information so if we can find a way to exploit it we can crack problems that have otherwise completely intractable but please keep in mind that it's not easy to find the right kind of algorithm that actually Explorer it's all that exponential computing power okay so there's only a few examples now so far and finally the quantum computers need not be a completely different object from what you're used to in fact people like myself and some other colleagues are building quantum computers using the same technology that's used to make the computers you have nowadays so we didn't finish I'm happy to take questions [Applause]

Channel: SibosTV

Views: 369,374

Rating: 4.9095154 out of 5

Keywords: Sibos, Sibos 2018 Sydney, Sibos conference, SWIFT Institute

Id: 7susESgnDv8

Length: 39min 14sec (2354 seconds)

Published: Wed Nov 21 2018