IonQ CEO Peter Chapman on the Future of Quantum, the Tech and the Company

hi this is pat moorhead with more insights and strategy and we are here with peter chapman ion cube president and ceo for another insider podcast how are you peter i'm doing great thanks matt thanks for coming on and as we talked about in the green room i know there's nothing more you would love to do than do a a podcast and get your message out and you do have an incredible story and a lot of people are very excited about that but for those who might not be too familiar with quantum computing out there uh can you maybe we can start off uh by talking about kind of what it is uh why it matters to not just technologically but you know to society and and the market so quantum computing is a whole new mode of computation um it's uh it's unlike today's uh computers in in kind of every uh regard um it's it's akin if you will to more how a gpu works next to a classical uh computer and why does it matter well it turns out today's classical computers even though they seem all powerful and they can do just about anything it turns out they can't and so there's a set of business problems which we need a much more powerful computer to be able to solve many of to be honest one of mankind's grand challenges and so we think quantum computing can be the solution to answer everything from you know maybe climate change to drug discovery to new batteries um you know literally many of the challenges which sit ahead of mankind today yeah it is exciting and you know i've been researching uh i guess we call it classical computing for over 30 years and uh you know when you look at the architectures and even moore's law we needed something different um and you know an interim solution appeared to be the accelerators like a gpu or an asic but it looks like the end you never want to call anything the end game but but quantum computing is is as close as you're going to get to the end game as possible it's maybe not for all problems but for a subset of problems they definitely seems like this would be a great solution yeah so uh you come from a family of futurists uh your father was a nasa astronaut which is cool not too many people can can say that uh how did you become interested in quantum computing and why did you decide to be the ceo of inq so um so i started in the computer business back in the mid 70s my dad brought home an early apple ii and i just fell in love with with you know these new uh computers and um very quickly went and started working at the mit ai lab under marvin minsky and you know i fell in love with with ai and i'm talking about the ai nut we that word has gotten you know dual use now there's machine learning and then there's kind of the ai and for movies i'm interested in the ai for movie part and so um but back then with only 8k of memory it became pretty obvious we weren't going to build you know those kinds of programs with that hardware um and so i kind of shelved that and got out of ai for many many years and when the cloud computing started to come into play i thought well now maybe i could have a system powerful enough to actually build it but then realize that's not that's really not powerful enough either and so what i really needed was a quantum computer to solve that problem so um i kind of joined the company because i actually need the quantum computer myself for my own kind of private hobby which is to build a an ai system and sure enough i do think that quantum now um will be the thing that breaks through kind of um that kind of ai system that we see in movies in the future so some people uh bird watch uh other people uh watch um a series on netflix and for fun you build uh ai systems so i think that that in itself uh says says a lot about you very very very very impressive and i think i told you this before you know whenever i walk into a room if everybody involved in computing in quantum computing i could pretty much say okay just want to let everybody know i i do not have a phd uh but i'm here to explain the magic that that you create to the uh rest uh rest of the world so uh there's a place for all of us but but very impressive so the potential for quantum uh is is massive it's huge and it's funny and sometimes you know people are like well what's the growth going to be right it's it's it's it's big enough to matter we don't have to have an exact number but it's it's massive but can you tell us a little bit about uh how inq was formed how was it started and also the team that you built because what i understand is that some of your team members come from the very inception of of things like uh trapped ion so let's brag a little bit here yep no um the two co-founders chris monroe and jung seng kim um have been in quantum since the very beginning in fact actually chris uh was at nist back in 1995 did the world's first ever quantum logic gate and that really kind of started this this whole thing off um and so um since then the two um their professors at both university of maryland and at duke have been working on quantum using government funding and i have looked at many different ways to solve this problem and decided that ion traps are probably the best solution kind of going forward back in 2015 uh chris and jung sing wrote a paper that um got picked up by a guy named harry weller who was running the east coast division of nea the huge venture firm and god knows why he was you know watching reading scientific papers but um it probably says a lot about harry and his success and and so he read chris's paper and junk sayings and he thought this read like a business plan and so um nea approached you know the two co-founders to put money in which is pretty unusual unto itself because usually it goes the other way you know you go chase the vcs so this was the vcs chasing us so that began became the beginning of ion q with the seed money that came in from nea and then very quickly they got google and amazon um nnea to come into the first round and um and then what was very cool is they put together a deal with the university of maryland and also with duke to exclusively license royalty-free the ip that the two professors had um developed over the last you know 15 years and going forward until 2026 so um so we kind of work in conjunction with the universities um and and obviously inq that's kind of the basis for for inq yeah that's a that's a cool cool story um some things come to mind you know like the a-team uh or you know these new superhero movies of people who you would know kind of coming together to uh create something sometimes there's friction uh but most of times they they do uh some pretty crazy stuff and i hope you hope you do that as well uh not the friction part but the writing big stuff yeah well luckily the um it turns out that uh that on chris's side he's the physicist and on jung sign uh jung seng's side he's the um engineer and so the two of them actually come together so there isn't much friction i think they they almost between the two of them make a make a whole person in some sense right so they they compliment each other in ways which is really cool so the symbiotic system i i understand that so uh that's kind of my wife and i so yeah so big news uh recently in that uh ion q quantum computers are now available on google cloud and essentially that makes you the only quantum computing provider on on the big three right now so google uh aws and azure and i mean i i feel like i've explained this in significance to people but for those who don't from your point of view and your customers what is the significance and the value of being available on all on the big three um well there's certainly there's there's kind of multiple ways to look at it first obviously is just ease of getting to our hardware um you probably know you know it's often could take quite a long time for bureaucracy to figure out how to open an account and you know become an approved vendor but everyone has between the big three everyone has an account with those guys so um so it's easy to piggyback on that um but the other aspect is um google is a you know there's today there are several software development toolkits for quantum and google has one and so in addition to being available on their cloud we got software support for them for their quantum languages to run on our hardware and so um we have wanted to make sure that everyone's tool kits doesn't matter what you know who it is actually runs on our hardware so that everyone there's going to no excuse not to not to be able to run on our hardware yeah and listen uh from my point of view and i cover multiple workloads on on the big three and and net net people want options right and and typically they have a primary and they have a secondary cloud cloud vendor and you know if if the one quantum option is your tertiary then it's just more work that you're going to have to do to link them up into your into your operations exactly and now at this point we've covered most of the cloud in the united states here with these three and now covered almost 100 of the software quantum software toolkits as well yeah so let's move to uh programming languages uh and and and dev kits here so uh you recently published uh this chart that i'm gonna i'm gonna flash i'm gonna flash up here yep um yeah you recently published this chart showing your compatibility with more quantum programming language and dev kits than any other company out there right now so what does this mean for customers uh especially in the light of of the announcement that you did just did with google cloud well it means that no matter what the tool kit that a developer is using for quantum it it will run on inq hardware and so we um we don't have a programming language that we're trying to push instead what we're doing is making sure that others quantum toolkits languages actually run on our hardware and it's easy for developers to to make that happen we want it to be a frictionless experience so we announced recently support for ibm's toolkit and also google's and that kind of you know completes now we have basically support by everyone in their various toolkits so no matter what uh toolkit a developer is using you can now run it on a high end queue how do how do kisket users get involved um kiskit users can either do one of two things they can either open an account directly with us and run directly or they could run it through google as an example and run it through the google cloud either way interesting in a way you're you're almost making uh quantum look easy uh i know it's not uh but when you remove objections uh out there it makes makes things a whole lot easier it's a lot of work don't get me wrong but it makes things move along quite nicely so um let's talk about differentiation uh my analyst firm warren sites and strategy and paul smith goodson we track a lot of different uh companies in this space and i'm curious how are you different how are you better than than your competitors um well first as we talked about everyone's software works directly on our hardware so that no one else can say that um so we have the best software support um we as you mentioned you know across all three cloud vendors no one else has that but in terms of the hardware itself we have the best quantum computers in the market today and um so there's you know what is it's needed to build a quantum computer and it's really two things you need to have good quality qubits and you need to be able to control them and so in our case we're using individual ions as a indiv as one cubit and um as such you know we didn't we didn't manufacture the atoms you know those come by mother nature every every atom is identical to every other atom in the universe as long as it's in the same atomic species and um so in that sense they're perfect so we say that these qubits natively are absolutely perfect which is true we have no manufacturing errors and so in and so we others when you look at it is they're trying to build a man-made qubit and so there's a problem they have a manufacturing problem because we just don't have technology to get down to manufacturing at the scale that's needed um in in terms of a man-made qubit but inq we're using we kind of say mother nature's cubit which is there the mother nature actually did the manufacturing the the next part is actually being able to control the cubit and these qubits are kind of i'll say kind of shy they do not like interacting with the environment in any way shape or form if the qubit is is bothered it will fall apart and it will stop doing computation so in our particular case what we're doing is we're taking the individual atoms the ions and we're suspending them in a vacuum and we've completely isolated them from from the environment and they're being levitated in a electromagnetic field on top of a little chip and so they're you know they're they're perfectly isolated from the world you know because of the vacuum um you know there's no they're not getting interacted with any other particles and the way that we program them instead of trying to attach a wire to them which would be a funny thing because i'm not quite sure how you attach a wire to an individual atom but instead what we do is we use a laser and and we use a photon and as you know photon light has a very you know very light touch to it so um so what's happening here is the lasers are coming in and hitting these ions and so it all in all has this extremely light touch and so what that means is that the um the system can there's a technical term kind of t coherence but it really means is they stay together so that they work so that you can actually run a quantum program if you look at some of the competition is they're building these they're using silicon and they're building a man-made qubit and if you actually you know an individual cubit might be 10 to the 14 atoms quite a few atoms if i actually looked at it as under a microscope you'd see it's fuzzy and it that looks like a shag carpet from the 1970s and and that's just showing you that we just don't have the manufacturing technology to lay down atoms one atom at a time in exactly the structure we need um and then because they're mounted onto a silicon substrate all the noise and everything that comes from that environment then comes into the cubits and they don't like it and then they tend to do deco here so what they have done on that side often is to try to lower the temperature down to zero degrees kelvin and on our side because we've got it in a vacuum we're running them at room temperature and so there's these huge differences in the architecture the bottom line at the end of it all is that the ion trap technology we're using has the best fidelity meaning it can run the most complicated circuits in the marketplace today and so we it's not uh very controversial ion traps kind of beat all the other technologies in terms of the fidelity um and that's important when we get to the next phase of of computing right which is right it is all about how do we get to error correction and yeah and i was going to ask like how many qubits do you have to get at before you have to start using error correction well um it really depends on on your error rate um so if you're if this error rate that's in these we call them noisy quantum computers if the um the error correction is is dependent on the error itself we use error correction today actually in the in this video between us is is the internet and those are all error corrected as well and so classical computing we use error correction all the time in quantum we need to do the same and the amount of error correction you need to do is dependent on your error so if you have a 95 percent error rate it really doesn't matter how many qubits you have because only three or four of them would actually be usable you know you could have um a hundred qubit chip and with that error rate three would be usable in a million qubit chip it wouldn't change the the usability of the chip it would still only be three so um and then what's worse is that if you have a high error like a 95 percent then maybe the the amount of error correction you might need to do an error correction here what you do in a quantum world is you use more physical qubits to make one better one and so the overhead for that might be a million to one you might need a million cubits to create one good qubit inq is the only company in the last year with duke and university of maryland to show an error corrected qubit working and we did it with only 13 uh an overhead of only 13 qubits and the reason for that is just simply um the error rates are so good that the overhead for the um you know for the error correction can be small so it would not be unusual for other technologies to have a thousand ten thousand a hundred thousand or even a million in needed in physical qubits just to create one good one and so we have that advantage in the second phase and because we're the only company that has actually done an error corrected qubit where it's actually shown to be better so that's kind of the next phase uh in kind of the maturity of the quantum market well that brings me to my next question uh which is assessing quantum performance i've been i've been on benchmark consortiums and you know my background is product management and and product marketing so i've done my share of bench marketing uh as well i've worked on industry standard benchmarks i've blown up benchmarks i've created benchmarks and i'm curious how do we benchmark today uh how do we know which one is is the best beyond the scientific explanation that you just uh gave and how should investors customers and developers think about which hardware uh to go with in lieu of having you know any is something industry standard that everybody's agreed on well we we definitely as an industry need to create a standardized benchmark which um is not run by the vendors you know is run by you know we need we need one of those um uh sticker pages you see on cars right it says this is the average gas mileage you're going to get with your quantum computer um and so today there is a few attempts at doing some sort of early benchmarking ibm has a a benchmark called quantum volume and it's basically trying to unlike the classical versions which are trying to tell you how quickly a processor works this benchmark is trying to tell you how big a quantum program you can actually run and because the larger the program you run the more commercially interesting it is and so um so this is kind of roughly but there's a lot to benchmarking um in terms of of kind of what's the right benchmark and you know and how to do it generally we've put up a calculator on our website which takes the ibm quantum volume and it it we we kind of um shrank it down we call them useful qubits it's based on the ibm quantum volume um but in quantum volume they exponentiated the result and instead what we did is we took the log base two of it and what it gets you is the number of usable qubits it tells you basically out of this comp out of this quantum computer for most applications how many of the qubits the the basis of the computation can be used in your quantum program and so um you know people can use that calculator and type in competitors numbers and they'll see what the number of usable qubits are and if you went and did that you'd see that inq you know wins fairly easily so you asked is this is an estimator and and i'm wondering uh why doesn't uh inq just run it and and issue something as opposed to uh estimating it or maybe you have and you just haven't published it yet um so on uh you know many people have published uh results you know they come out to be honest every month somebody comes out in and compares our systems to the competitors we don't usually have to do it because there's lots of researchers and customers who do that all the time and every time they do we come out on top and so um we don't uh you know we don't feel the need for that but in general too since most of the vendors are not um they're writing papers about systems which are sitting in labs we would much rather have customers and independent third parties to be the one that verified these things right yes yeah that that is better uh i do like the drag race uh the drag racing because it gets more people interested right and we saw this with classical i mean we see this with pcs i mean today i mean uh a vendor in classical computing uh uh announced that they won 14 top benchmarks with their new systems based on spec right so it keeps getting people excited you know we've got the top 500 super computers so i'm hoping we get there uh ultimately um so uh the definition of algorithmic qubits does absolutely make sense as as a way to compare quantum computers but i'm wondering uh i mean there are other features that distinguish inq's architecture and approach correct that is very much so um so the there's um how we we uh program these quantum computers as i mentioned is just a set of atoms which are sitting in a vacuum chamber and then we wire them up if you will using light using a laser and so what our compiler does is it takes a quantum circuit and it kind of if you will builds the perfect chip for that particular program if you look at what our competition's doing you know often what they're doing is is they're building a chip at and they're designing at design time what the kind of least common denominator would be for all possible programs in the future and so or they will have to get to a point where they'll put out a custom chip for every application and so what's cool about our technology resting is is basically a compile time it's get you're getting a custom chip it's not literally but you're we've wired it up in a way that's the most um uh the best for your particular application and so that's really cool by the way that's the best explanation i have ever heard of of that feature and um i'm not i'm not doubting that it's true but i'll just say that that you know you get this going i mean that that could be that could be a game changer in in itself um the classical computer folks um are still figuring out how to create custom chips but they have to do it in the package and it took them 40 years to get there with 3d packaging and tsv's uh but the notion of customization sounds sounds pretty awesome it is it's very cool and one of the neat things about it is you can apply error correction in the circuit only where it's needed right because you know this error correction is kind of expensive right so you don't want to use up your presser's qubits when you don't need to so let's just use it in the places we need to in the circuit and nowhere else and so the compiler can really optimize the hell out of this chip for one particular application wow i love it i love it that's cool i'm gonna uh everybody we should just stew on this uh for a while um but hey let's jump into the ion cue as an investment but before we go into this i want to remind everybody this show is for educational information purposes only uh check out your financial specialist uh if you wanted investment advice now with that said let me jump uh into that into this so you're you're jumping into this uh head first with your spac and your pipe transaction a company's going to be listed new york stock exchange under the ion queue badge so where does ion q go uh here from here as a public company why would somebody want to invest now versus let's say uh uh in in in the future i mean we've got some of these pundits saying that uh um uh uh quantum computing is uninvestable i heard a uh a podcaster and a vc say that the other day yep um just on just interesting a couple of things when i first joined the company two years ago there was a major article by a professor out of the europe who said quantum computing would never work and um and i i think actually wilbur and orville the week that they flew at kitty hawk there was also somebody who said flying would never work too it was i read that i i read that autobio sorry it wasn't an autobiography but an incredible uh book about that and i remember that and i think it was uh the folks over in france that made uh that made that comment made that comment yes so um what's interesting is um in the last year we've put our systems out in the cloud i don't hear from anything from those people anymore because you know for a dollar and a credit card and one of these accounts you can go write your first quantum program so you know we've kind of put the doubters um behind us so the question as to why today to invest um that's really kind of the purpose of specs right you could you know we could wait for the company to go to a traditional ipo um but by that time all the not all but a lot of the value has already been created and you know you're missing out on it and so the question is there is definitely higher risk in hopefully a higher return here but the question is is do you want to capture that upside as we get to being kind of you know the a traditional ipo company and missed that upside and what is that upside between now and in kind of that future is um uh every generation of hardware that we come out is getting closer about 72 qubits you can take on the world's largest supercomputer or take on the cloud itself for certain applications and the world's largest supercomputer they tell me runs uh it takes the power of hoover dam to run but our quantum computer will be able to compete and it runs off a standard wall socket and so as we make progress on the technical progress to that um each one will be a validation and an opportunity for the stock price to go up and in addition um you know we're we're now working with customers on various applications and each one of those is a potential for a breakthrough and so it is definitely early days it's kind of like um the beginnings you know the mid 70s in the computer business we're still waiting for bill gates to come along and produce quantum basic and we're waiting for dan brooklyn to come along with quantum the quantum spreadsheet so it's early um but as soon as those things start to appear then the thing's going to take off and so the question is is would you who would have rather invested in intel in the 1970s or waiting you know much much later i would argue you probably should have invested early so um yeah peter and i saw it pick off a couple other uh reactions uh in in the industry uh that you know in the past month that i i thought were fascinating as well and i think that's i think that's good and that's good for the market uh definitely competition is always good and it's also um showing that this level of financial support um is also good for the quantum industry so i think yeah for the entire year well and you're partnering with uh dmyi right that is correct yep so the uh today if investors wish to invest in the company they can do so by um acquiring dmy i stock which when we get to the public listing we'll switch over to inq the inq ticker symbol which i'm very happy the company um very early on decided to choose a company name with only four letters so we didn't have to run through the usual exercise of choosing ticker symbols yes and by the way thank you for doing it like like you did drives me crazy uh and i think for especially for retail investors it just makes it a lot easier to define the company when they're trying to buy or sell agreed so uh uh peter uh i could sit here for another hour and do this and i think our listeners would love it too but uh uh we always try to keep it short here on the more insights and strategy insider podcast i just want to thank you for for coming on it was really a fascinating discussion and every time i talk about it uh i i get educated uh and give the ability to to do better analysis and i i love what inq is bringing to the table super super exciting um pat it's it's always fun i look forward to doing it again in the future so absolutely thank you so much and then congratulations again on your on your google uh uh win there as well i just throw that in there thank you um thank you we appreciate it yeah so this is pat moorhead with more insights and strategy uh closing out another more insights and strategy insider thank you for coming if you like what you've heard hit the subscribe button up there if you have any suggestions uh hit me up on twitter it's the easiest way but with that i'm going to take us out of here have a great week

Channel: Moor Insights & Strategy

Views: 31,556

Keywords: ionq, dmyi, quantum

Id: ttkmk7mdExM

Length: 35min 21sec (2121 seconds)

Published: Tue Jul 13 2021