Episode #8: Quantum Computing - This is Why It's Important

Computers now have primary control of critical vehicle functions.

Welcome to the IBM Keep It Simple Technology Podcast, where each episode we discuss technology but in a simple way. We keep the explanations simple, but what we spend our time on is emphasizing what this technology means to you and how you can best utilize this technology for you and your business. Your host, as always, Thomas Anglero.

Welcome to the IBM Keep It Simple Technology Podcast. And I am your host, Thomas Anglero, Director of Innovation for IBM. And today we have a special guest. We're going to focus on quantum computing today. And our guest, I mean, the guy is crazy. He's coming to you from a shack in the middle of Norway.

I'm not lying.

Is Lars Nordbrin, who is our senior architect and an IBM Q ambassador. IBM Q is IBM's quantum computer. Lars, are you with us?

I am. Thank you, Thomas. Hello.

You got to tell people about your shack, Lars. Just set the framework because we're all in lockdown today. Tuesday, March 31st, 2020. We're all in lockdown. And Lars, you're locked down in a shack with what next to you? This is hysterical.

Well, basically, this is my man cave, so I need to have some isolation also from my 3 girls who are doing their homework in the house. I have a shack isolated from the house. I have my guitar, my CD collection, my coffee machine in an arm's length distance, and I also have my brewery here. Lars, I know. I mean, what did people do before they have internet in their brewery? I don't know.

There you go. There we go. Spoken like a true nerd. Oh my God, Lars, great. Don't drink while talking on a podcast, okay? Keep it to after the podcast. Lars, there's a whole bunch of people listening on this podcast right now, and a lot of them are either in their man cave or hoping to God they'll have a man cave in the next 5 minutes to get away from the chaos behind them, um, or on top of their lap.

I remember those days.

And let's start off directly. When it comes to the world of quantum computing, this is a word, a subject that has been talked about, hyped up massively, absolutely massively. So why don't we— why don't you break down really simply What is a quantum computer compared to a normal computer? Go for it.

Okay. Yeah. To start with the normal computer and the terms that people are acquainted with, normal computers are registering all their information in zeros and ones. This is kind of the usual setup that people are acquainted to. The quantum computers are quite different. So they run on what we call qubits. Not bits. Qubits is a series of units that are using quantum mechanics principles. So they are not storing and computing in zeros and ones. They are computing in a combination of zeros and ones. And then in addition to that, there are also other characteristics of quantum mechanics that comes into place where each of these units are actually working with each other. So they work together also. So there's, first of all, a lot of more information to store into those qubits. And there's also different things to do with them than with ordinary bits. So it's not kind of an equivalent, just a scaled-up equivalent to a classic computer. We need to think completely different on how we are to solve the questions and the problems we are going to do.

Okay, now let's keep it simple. And let's break down zeros and ones in quidbits. I love the word quidbits. You should call that for your new beer, Quidbit.

Qubits. Sorry.

That's okay. So here we go. Let's give it— let's break down the zeros and ones, because I always think that's so such a nerdy statement until you understand how simple it actually is. So in the computing world, 0 and 1, when you program, I like to give the world's simplest answer example. And that is if I say the word hello, right? For a computer to program the word hello, which is spelled H-E-L-L-O, the letter H is actually a defined code of maybe 001. That's the letter H in computer code. And then the letter E is maybe 0, 1, 1. And then L is another combination of something. So when he talks to— when Lars says zeros and ones, everything you do when you spell a word or do something on the computer, every minute action is either a zero and a one or a combination of a zero and a one. So that's what he means by zeros and ones. And on a supercomputer— I'm sorry, on a quantum computer— sorry, on a quantum computer that this is always very foggy for me because it's so different because it's quantum mechanics. The computer is not just at a state, things are not just zeros or ones, they could be zero and ones at the same time. Am I right, Lars?

Yeah, I would rather call it a combination of a zero and a one. So yes. Yeah, exactly.

And this is why quantum computers are so amazing. Prior to quantum computers, everything was very, you know, 1 1 2. That's the way we see things. But in the quantum world, things don't happen that way. They've proven that an element or an electron can be at two different places physically, geographically in the universe at the same time. That doesn't make sense in our normal world, but in the quantum world, things are different. So in the quantum world, when they compute with zeros and ones that are not just zeros and ones, but they're zeros and ones at the same time, you have tremendous amount of advantages and benefits. And this is where a quantum computer leverages that. Lars, in this quantum world, I think— I hope now everybody listening understands that we're dealing where we have to think in a quantum level of things where things aren't basic. What are the benefits that come from using a quantum computer versus a normal computer working in the quantum world.

When you scale up these computers, you are able to combine several, call it values or problems, and combine them in a much larger extent than with the classical computers. So you are actually testing several things at the same time. That's how you are actually using that quantum mechanic effect that you are testing a combination of different values at the same time.

That might not— Hold on, Lars. Let me just clarify. What do you mean by— And because with a normal PC, you can run a program and it's parallel processing. So compare that to the quantum machine. What's the difference?

Well, let me take an example. If you have 4 cards on the table and you know that one of these cards is a queen, you're turning them upside down, and the way to find the queen is to flip each card. If you do that 100 times, you will have an average of 2 tries to find the queen. That's how classical computers do it. But the quantum computer can test several of these possibilities at the same time. So they will always pull the queen the first time.

So it's true parallel?

Yeah, it's more than parallel, but yes, you could use that term.

Yeah, I'm trying to use that term because the only term I could think of in our— outside of the quantum world, our normal world. Wow. Okay.

Yes. So it's, it's, it's in an unstructured search, you'll try to, to, to test all the different possibilities while you in the quantum context can, can do several tests at the same time.

Exactly. So in the normal world, it tests, it tries, it does trial number 1, pass, fail, fail, trial number 2, pass, fail, fail. And it keeps on going until it passes. And in the quantum world, there is no trial 1 or trial 2. It just does it all at once.

Yeah. Or more efficient.

Yes.

Yeah.

And more efficient.

Wow.

That's okay. Keep going. That was—

Yeah. And the joy of this, and I think that's one of the other things that we need to understand. When you have a set of bits in a classic machine or you have a set of transistors in the in the chip of a computer. If you are, if you want to double the calculation power of this machine, you need to double the number of transistors in that chip. While in a quantum computer, if you have a set of qubits, if you want to double the calculation power of that quantum machine, you just add one qubit because it has an exponential growth for every qubit you add to that system.

Okay, now let's clarify, what is a qubit? Is that a physical thing?

Yes, it's a physical thing. Uh, so a qubit is emulating, uh, what— and this is not able to do simple. And I, and I, if I would like to quote one of the, the founding fathers of quantum computing, Richard Feynman, he said, those who understand understand or think they understand quantum mechanics have not understood quantum mechanics. So just have that in mind.

But good luck with this podcast episode.

But an atom or the lowest units of our physical matter, they have different what we call states, quantum states. And that is actually what we are doing with the qubit is to try to emulate that quantum state. So an atom can have a quantum state, a photon, which is kind of the basics of the light, has quantum states. Several of these particles have quantum states. And a qubit is, It's a setup. We are emulating that quantum state, and quantum state is a 0 or 1 or a combination of 0 and 1s. So what we do when we, when we make these qubit chips is that we are trying to construct that quantum state with electronics.

Exactly. And is a qubit chip Is one qubit one qubit chip?

No, we are, we are adding more and more qubits into one single wafer. So in our largest quantum computers to date, we have 53 working qubits on one wafer.

Exactly. So that's somewhat similar to the standard CPU model where they're packing in more and more transistors per CPU wafer, correct?

That's correct. That's correct. So we've seen a massive growth just over the last 2, 3 years from just 5 qubits to 20 qubits and now recently 53 qubits. And that's one of the races actually in the market also to see how many qubits you are able to put on the wafer and how stable you can make them. How noise suppressing we are able to make them and so on. So it's not the perfect technology for now. So it's very seldom that you see we are opening our research programs at such an early state that we're doing with quantum computing. So that's why it's also been discussed more in a lab term than actually in computing terms nowadays.

In the quantum world, when you add another qubit to a wafer, that is not the same as adding just one more transistor in the CPU world. Because as you said very clearly and very well, Lars, that when you add a qubit, that's an exponential growth. Where in the CPU world, our normal world, you add a transistor, that's not exponential growth. So it's nothing at all the same. So everybody listening should understand that This is not same same, right? Because a lot of people saying, well, I have a Pentium computer. I guess I'm dating myself. There aren't Pentium computers anymore. But, you know, and it's got this or that or whatever. It's not the same. We're talking about exponential growth when you add one more qubit, sorry, versus the whole transistor model, which is totally different. Lars, this is— what are the benefits to using a quantum computer? Who was this? To start off simpler, who is the market for using quantum computers today? And then as a business leader who has a business and going, wow, I would love to run some analysis using a quantum computer because it could analyze everything in every scenario possible all in one second. This is fantastic. So who's the market for quantum computers? Today, who's using it, and then what businesses should start looking at quantum computing because they could benefit from it today?

So, so at this time, we are actually in, in several research programs with different actors of different industries to find use cases for quantum computers. So, so at the current state, you cannot do anything with existing quantum computers that new cannot do with classic computers. So that's very important to emphasize. So this is the nature of the research program is that we see when can we see a benefit in terms of 5 years, 10 years, 15 years, and so on. So today we have no benefits for anyone. So it is a research program, so to speak, just to make that clear. But the areas where we see some very interesting things happening is obviously going back to quantum mechanics and seeing, doing chemistry models with quantum computers. So in basics, chemistry is quantum mechanics when you are going into the minor details on that micro perspective. And to make precise models, you would need to have quantum mechanic-based models. So that's one of the major first purposes of a quantum computer is to emulate chemistry. So that could be interesting for process industries where you are optimizing certain processes and making it more efficient to generate new compounds, for instance. And I mean, there's a lot of different purposes within chemistry.

Yeah, I mean, right now it jumps out with the whole COVID-19, being able to identify new potential cures for this. That's definitely an area where quantum can help.

Yeah, so that's in drug discovery, for instance, but not all. Also that, also in how to understand how these proteins work and so on. It's very complex, so you are probably not seeing that much purpose of quantum computers within the first couple of years, but it might be interesting in the long term, perspective, yes.

Okay, and then companies today— that was great, good. Continue, please, don't interrupt you, go for it.

Yeah, and then we have use cases that are not linked to quantum chemistry at all, but is more linked to the possibility to test several combinations at the same time. So we have a discipline within mathematics called optimization, which is also quite fun when you look at how incredibly complex a lot of the problems that we are very comfortably having around us, and that we think that we are actually doing the most optimized way of practice, but we really don't know. And in optimization, a quantum computer might be able to do things a lot more efficient than the classic computer. So, and I've used that example a couple of times about how incredible complex problems really are when you look at it mathematically. If you, for instance, have a table with 10 people around it, how many combinations of seatings do you have on a table with only 10 people. You know, with 2 people around the table, it's 2 combinations, you can switch chairs, that's all. But when you have 10 people around the same table, it's actually more than 3 million combinations of seatings just around that table with 10 people and add the 11th, then you have 40 million combinations. So, so these, if you try to find the optimum way of seatings, you need to test all those 3.5 million combinations of people. So that's an optimization problem. And there's lots of those optimization problems. And that's an area where quantum computer might be efficient. And kind of a spinoff from that also, which is another interesting area is, is financing and, and how to modulate the finance, uh, dynamics, uh, on a, on a macro scale. If you see, uh, how is the markets impacting by this and this and this and different scenarios and this different scenario testing. So the financial market does not have a lab. Chemistry have a lab, electronics have a lab, computers have a lab, but finance does not have a lab. So if you're able to make better models of how you are, how your financial industry is acting on different, different scenarios, that will be extremely efficient.

Absolutely. This, this to me, I'm getting more and more excited as I listen to you because it sounds like we've been looking at living in the caveman days compared to where quantum and its efficiencies will drive us, right? Being able to look at millions, billions of different scenarios and possibilities and get such, or closer to, more of a definitive answer. Um, Lars, uh, what about programming? I think this is something that nobody ever asks, and I think people need to understand that. Can, if someone was interested in doing, getting into the quantum computing, is normal programming skills enough to program in the quantum world? If they say, yeah, let's get into it, let's start a project, um, or do you need to have a special skill set to get into the quantum world and start programming for quantum computers?

Today you'll certainly have a very different skill set, um, but if you take it in a bit of a retrospect also, when you entered the computing business in the '50s, you would probably know a lot about transistor characteristics and Boolean algebra. While today you are just pulling strings between dots in the graphic user interface. So we are at the '50s level on quantum computers today. So you need to understand the quantum mechanics, you need to understand the qubits and the superpositions and the combinations of zeros and ones and, you know, all that entanglement and all that fancy words from the quantum mechanics. But we are currently, uh, as speaking and as an open community developing frameworks for programming, uh, also on high level, so that the, the, the future of quantum computing might be on a high-level perspective. You have your use case and you're programming it with it, with, with the sensible set of, of parameters, and you get the result. But, but the core of it, and that's where we are actually doing work today and optimizing it today. You need to understand how you are abstracting your use case into wave functions and quantum mechanic principles in order to make the quantum computer calculate it. And it's also important to emphasize that this is not something you do overnight. This is Exactly. This is an area where I would hope that all universities and all industries are starting to investigate and invest their peoples and skills into it because this is something we need to do together. We need to develop it together in order to be prepared when the hardware is strong enough. So it's a separate activity actually to learn the skills of programming, understanding the core fundamentals, and also investigate what kind of use cases can be implemented, what cannot be implemented, where is this good for, what is it not good for? Because there's so many use cases that quantum computers will be really bad at doing. So it's a very small subset that we see dramatic improvements, but there's so many things that it will not give the improvement. So we need to find those areas. And that's a lot of the research programs now is to understand which of them, which of the use cases that actually will benefit from a quantum computer and where you will benefit most.

Lars, you left off before when you were saying that you needed, you know, the list of things you needed to have to understand the quantum computer, you left off a couple of critical things. And let me just fill in the blank. You need to have a shack, you need to have a brewery within arm's distance of yourself. Am I right? There's some of the secrets to quantum computing.

You need to spend some time for yourself, at least. Yeah. Now, I think that when it comes to the basic science, a lot of these, these poor science Scientists have spent a lot of time by themselves thinking and doing things in an individual manner. And it's not always easy to understand which of the scientists and which of the inventions and new principles that actually will come to some use at the time. But yes, I'm starting off with Brüri and the CD collection, and I'll take it from there. CD collection. It's no, no offline. There's only offline music in my shack. And actually, I'm compiling a list of crucial albums that you, you really don't find by stream, only on, on physical media.

Look, I love you, Lars. You're great, man. Listen, everybody's locked up in quarantine and stuff, and as soon as the quarantine is over, party at Lars's Shack. We'll give out an address on the website of the podcast. Lars, uh, let's wrap up this episode. You've been so informative and you have kept this simple, and I mean, this is a very difficult subject. Any last thing you'd like to share?

Um, well, I'm— yes, actually, uh, I'm, I'm opening this for, for, for people to to do their own exploration. So if you Google up IBM Q on the internet, you will find yourself a way into our open community where everyone, and I mean everyone, can be member, test it. We have our quantum computers available online on the internet for everyone to test. There's more than 100,000 people currently that are members of IBM Q Experience, and that's really the way to start for everyone. And, and as I said just a few minutes ago, we don't know which of the inventions or, or new ideas that actually will crack and be the next big thing. So, so it's open for everyone.

Lars, that is a fantastic way to send us home. I thank you very much. Lars Nordbrin, Senior Architect, IBM Q Ambassador. Thank you for your time. Ladies and gentlemen, you know, Lars gives us great advice, you know, experiment and play. And as he said, these are the early, early days of quantum computing. But if all you remember the early early days of the dot-com days, we all wish we would have taken advantage and did that thing, and then we'd have been so much better off today. This is another point in time in the technology world to get involved, get excited, and see what comes from it. Lars, thank you. Everybody out there, thank you. Like the podcast, share with your friends, families, and stay healthy. Stay safe until the next time. Thank you very much.