Google unveils 'mind-boggling' quantum computing chip | BBC News

now the latest development from Google that it describes as a state-of-the-art quantum computer chip the company claims Willow Leap Frogs over these supercomputers that we're used to we used to consider the world's fastest and it takes five minutes for them to solve a problem that would currently take supercomputers 10 septian years to complete well Professor Alan Woodward is a computer scientist at s University um septian before I ask you what how much is that I know it's a lot but first of all I'm going to start with uh with with you know the very basic

question if you can explain what a a quantum computer chip is right that's the million dollar question the um in conventional computers the sort the laptop I've got in front of me the the unit of information that's used for processing is called a bit most people have heard of a bit and it has two possible values zero and one and typically the way they're implemented on a chip is with a transistor which is a fancy switch so the switch is either on or off so it's either zero or one in Quantum Computing you use what's

called a Quantum bit and the quantum bit can also be zero or one but if you can get it if you can coax it into this very special State um it's a complex combination of zero and one um it's some people argue that it's all possible values at the same time between zero and one um and if you get lots of those working together then you you can actually turns out there's some algorithms you can do which are much faster in that sort of State with a cubbit than you could ever do on a conventional

computer um I'm slightly hesitant to say that this is a the comparison with the supercomputer is actually really a valid one because the algorithm they implemented is really all about um it relies on quantum physics and so the first thing the supercomputer would have to do is simulate the quantum computer um or the quantum circuit so I'm not sure it's a fair comparison in some ways but the more interesting part about what what they've done what Google have done in this paper they they've put out is what's called error correction which is a very boring

subject most people will never have heard of it but we're all relying on it I'm using it right now all our modern Electronics um have errors in them all the time and if you're not to end up with just lots of noise but actually a signal you can use you have to do what's called error correction and what they've worked out is how to build um a chip with 100 over 100 cubits on it such that they've reduced the error rate below a magical threshold that everybody's been trying to get to for a long long

time time um and that's that's the big breakthrough if you like so it's a I hesitate to say it's a breakthrough I think it's probably more of a major Milestone and it's certainly very encouraging for anybody that's working on quantum computers because it means that that that what we can do is we now know we can scale these things up we can add more and more cubits to the chips without necessarily just incurring a a consequentially larger number of errors and that was really what tipped modern Computing over into sort of the race we saw

in the last 60 years where every 18 months the power doubled um and that's that's what this holds out as a potential okay well I'm I'm still grappling with the idea of it being being n and one at the same time but moving on what are the practical uh benefits of this potentially well actually it goes back to why quantum computers were first um uh first moted by Richard fman back in the 80s which is physicists were trying to use classical computers to model things in a quantum state so when you go down to the

microscopic level Quantum effects mean that you can no longer really predict and model things in the same way that you can in in our the world that we experience every day um and so they decided well why don't we actually make a computer out of something that's Quantum um and actually has the quantum effects built in um and that's actually what it turns out it's best at is modeling materials right down at that microscopic level so if you can imagine something like a drug company that's trying to um model uh various molecules and how they

interact the quantum computers hold out the hope that they can do that much faster than they can with a conventional computer and they can run through all sorts of combinations that would take them potentially years on a conventional computer and do it hopefully in minutes on on a quantum computer so potential um health health uses there what about the risks the risk uh the risk is one of the reasons people became very interested in Quantum Computing including myself I have have to say um was that back in 1996 um it turned out that one algorithm

that could be run was developed by complete genius called Peter Shaw um and what it does is most modern encryption it relies on certain mathematics that are very difficult to do one way um but very complicated to undo and find out the answer so for example if I ask you what's three * 5 I'm sure you'll quickly say 15 but if I were to ask you what two prime numbers multiply together make 15 it takes slightly longer to work that out quantum computers and shaes algorithm mean that it it's equally fast both ways so you

you end up undoing or making a lot of modern what's called public key encryption um basically useless it can be undone very quickly but it's not to panic because um we recognized this quite a few years ago um and indeed now people are being encouraged to move on to what's called postquantum crypto systems and to show that quantum computers are not just conventional computers sped up but they're they're very good at certain algorithms what we've come up with now is crypto schemes that are not don't have this and quantum computer will take just as long

um to crack it as a conventional computer if not longer so there's kind of risks but there's a solution there that is good to know it's absolutely fascinating which we had longer but Professor Alan Woodward thank you so much for explaining that so clearly great to talk to you this is BBC News