“Quantum computers are not unlike F1 race cars [in that] they are extremely expensive, very, very fast at certain circuits, but pretty much useless for bringing milk home from the store, carrying your kids around, or picking up a sheet of plywood.”
—Duncan Stewart, director of Technology, Media and Telecommunications Research, Deloitte Canada
Tanya Ott: When I say quantum, you say …
Male voice actor: Huh?
Tanya Ott: When I saw quantum, you say …
Male voice actor: Ummmm … What?
Tanya Ott: I’m Tanya Ott and today on the Press Room, we’ll break down quantum computing and tell you what you need to know for 2019.
So … just what is quantum computing? I posed that question to Duncan Stewart. He’s director of research for Deloitte Canada in the areas of technology, media and telecommunications. He’s also a co-author of this year’s global TMT Predictions report.
Duncan Stewart: The critical thing about our prediction on quantum computers is you don’t need to understand quantum computing. It’s an incredibly complex topic involving superposition and entanglement, and people with PhDs in the field cheerfully admit they don’t understand quantum computing. You don’t need to do that to understand our prediction. What we’re looking at is that quantum computers are a new kind of computer that do special things. They’re going to cost a lot of money. It’s going to take a long time to develop them. And what we do is we end up sizing the market—how big will this market be, how long does it take to get there, and what will it be used for. The takeaway from the prediction is, in fact, that most people don’t need to worry about quantum computers, but they are going to be seeing news about them probably this year or next, and when that news comes, I hope our prediction allows people to place that into context.
Tanya Ott: Okay, so without going into detail about what they are and the technology behind it, what kind of opportunities does quantum computing open up for businesses?
Duncan Stewart: There is going to be a thing that occurs—maybe this year, maybe next, maybe never, but most likely in the next year or two—called quantum supremacy. When that happens, that’s going to mark the first time that a quantum computer is able to perform a specific calculation using a practical amount of time and resources better than a classical computer. And I make predictions for a living and I predict right now it’s going to be front page news and everybody’s going to be talking about it. But what I want to do is place that into context.
The key part to what I just said was “that specific computational problem.” Quantum computers are not unlike F1 race cars [in that] they are extremely expensive, very, very fast at certain circuits, but pretty much useless for bringing milk home from the store, carrying your kids around, or picking up a sheet of plywood. In that way, quantum supremacy will demonstrate that for certain kinds of tasks—and I’ll talk more about those in the second—quantum computers will be better if you have the money and the willingness to pay for them. But for most of the computational tasks that we’re looking at around the world, both this year, in the 2020s, 2030s, and maybe even in 2040s, classical computers—the old-fashioned chips that we’re used to using—will still be the workhorses. They will be what almost everybody uses for almost all tasks.
Tanya Ott: OK, so you promised to tell us what some of those tasks are that they would be useful for particularly, I would imagine, in a business context. So, what would they be useful for?
Duncan Stewart: Even there, it’s kind of business-slash-academic. So weather modeling and advancing very, very complex weather simulations would be one. Nuclear physics—modeling the explosions of nuclear weapons without actually setting them off—would be something that quantum computers would be good for. A lot of detailed chemistry and biology. When we look at how molecules join together, that is a quantum process and the quantum computers of the 2020s—which we call NISQ or noisy intermediate stage quantum devices—will be actually useful for modeling the behavior of single atoms or proteins. And that’s an important business going forward. That will probably be worth hundreds of millions of dollars of quantum computing in the 2020s. Longer term, there are some interesting implications in logistics, in cryptography, and machine learning. So, all of that is sort of more longer term, but potentially disruptive and powerful.
Tanya Ott: That longtail there—what do you see? How is it going to be applied, for instance, to supply chain management or logistics?
Duncan Stewart: It’s really difficult to know because we don’t actually ... remember I said that whole quantum supremacy thing? Ah, that hasn’t actually happened yet. Why? Because we haven’t built a quantum machine that has enough computing bits—those are called quantum bits or qubits. Right now, the world’s largest, verified, tested, publicly-announced quantum machine has 20 quantum bits and we don’t even expect these devices to be useful until they reach 80, 90, 100 quantum bits. And they don’t become broadly useful until they reach millions of quantum bits, which is quite some way off in the future.
What will they be used for in logistics? We’re not entirely sure yet. Some optimizations some of the time. Is it the sort of thing that every company will have their own quantum computer? Almost certainly not. Instead, there’ll probably be a few quantum computers in the 2020s which will be accessed via the cloud and that's how people will tap into that potential. But to give you an idea: A very, very large logistics company might use quantum computing once a month, once maybe a week, but they will have thousands and millions of other computers that will be working on regular problems most of the time.
Tanya Ott: One of the other things I would imagine that quantum computing opens up—and again way in the future since this is not something that’s going to be really rolled out in any meaningful way for perhaps even decades or longer—would be that, as you can make calculations more quickly and do these computer processes more quickly, it opens up everyone to more cybersecurity risks. Have you spent some time thinking about that one?
Duncan Stewart: Yeah, that’s a critical aspect. One point I just want to go back on—you used the phrase: “They can perform calculations more quickly.” They can only perform certain calculations more quickly. For the average calculation with adding up and multiplying and subtracting and all that kind of stuff, quantum computers are not better than classical. They’re actually worse. They’re slower, more expensive, and less reliable. It’s only certain very, very narrow mathematical problems that quantum computers are better at. Interestingly, one of those is everybody’s favorite task, which is finding the prime number of factors of extremely large numbers. I mean, you and I do that every weekend.
Tanya Ott: Of course we do!
Duncan Stewart: But classical computers really—and this is a technical term—classical computers suck at that. It’s very hard for them to find the prime factors of very large numbers. And on that computational challenge, the entire edifice of public key encryption—which is what almost everybody in the world uses for their encryption—is based on the fact that factoring very large numbers are very hard for classical computers. Quantum computers are freakishly good at that. It has been predicted—not by me but by the National Institutes of Standards and Technology in the US—that the development of a sufficiently large quantum computer renders all existing public key encryption insecure.
So, one of the things that people want to think about is not so much that people are able to read your messages now, but it’s called a harvest and decrypt attack where bad people around the world can store traffic now, and if that traffic—which could be government secrets, intellectual property, or medical records—they keep that data for 10 or 20 or 30 years and when those quantum computers are built they harvest and decrypt then. So, in other words, if we want to start thinking about quantum-safe, quantum-resistant, if we want to start making our documents secure out into the future, we need to start doing it now. Bluntly, we actually needed to start doing it yesterday, but now is better than waiting 20 years.
Tanya Ott: One of the things I love about talking to you, Duncan, is that you use those really high-tech terms like “sucks.” But you also make it really real. And I want to transition us into our other topic of the day which is—some would call it additive manufacturing, regular people might call it 3D printing. And it is something that has been used for everything from making like a bobblehead of my husband or something like that to replacing parts, making parts more easy to get for businesses so that they don’t have to stock up and have warehouses full of things. They can make them quickly. So, what is your prediction for 2019 when it comes to 3D printing or additive manufacturing?
Duncan Stewart: There was a lot of hype around 3D printing around 2012, 2013. It was the whole factory in every home, the home market, plus even at the enterprise level, there was a lot of hype around especially final part manufacturing. Not so much the factory in every home, but the factory in every factory, and you don’t need a warehouse anymore, and it fundamentally transforms the entire logistics and supply chain. That did not happen. 3D printing got way ahead of itself and the growth rates in 2013, 14, 15 were quite low. The industry, it still grew, but only a few percent per year. Since then—and this is the core to our prediction—we are seeing the industry explode upwards. Growth is now double digit—almost 13 percent some years. And we expect that to continue to the point where the 3D printing machine market—just the large 3D printers that are used for enterprises and made by the big leaders in the field—we’re predicting that’s going to be a US$3 billion a year industry by 2020, which is great growth and obviously a critical size.
Tanya Ott: So what was driving the kind of laggard approach in the beginning few years and now this major explosion? Is it just that people finally figured out what it was or what the business case was?
Duncan Stewart: No, not at all. It’s really simple. The machines got better. They got bigger. They’re faster. They work in more materials. It just that, back in 2014, they were not quite yet ready for prime time and now they are. People are using them still a lot for prototyping and sometimes for intermediate parts manufacturing—making a tool, a jig, a die, a mold, a cast—but increasingly for final parts. But we do need to be super cautious right here. The global manufacturing sector is around US$12 trillion a year and the machines that are used to make all of that stuff are close to half a billion dollars in those various kinds of machines.
So, put in that context—do 3D printers grow? Absolutely. They’re growing much faster than any other part. But are they a big part of this manufacturing sector yet? No. Three billion compared to 500 billion of other machines—it’s still a tiny, tiny fraction of final parts that are made with 3D printers. It does happen occasionally, but that idea that 3D printers are allowing companies to get rid of warehouses and supply chains has not happened in 2019, will not happen in 2019, and it won’t happen in the 2020s either. 3D printers are really good for making certain final parts, but much of the time if you want to make part in large volumes, other technologies are usually the preferred solution and will be even in the future. 3D printing matters but it doesn’t take over.
Tanya Ott: Are there sectors in which 3D printing matters more than other sectors?
Duncan Stewart: Ooh, what a good question! Absolutely. Obviously, people are using 3D printers in medical all the time. You want to make for a premature baby a piece of bone or hip or whatever it is. That’s not something that you have in the giant [catalogs] from the large medical suppliers, so you print out your own. Lots and lots of medical applications. A little bit in fashion. Obviously, in fashion we like customizing. We like having things for just ourselves. Toys, the bobbleheads, that’s a real thing. There are people who make those.
Probably the more exciting ones would be in things like aviation. Aviation has a number of complex parts where you want to produce something in very small unit volumes, but [of] very high value. That’s an example where we are seeing 3D printed, FAA-approved parts being introduced into manufacturing. That said, when you look at a jumbo jet, 99.99 percent of the various parts that are on that jet are still made the old-fashioned way. 3D printing is used mainly for special manufacturing.
Tanya Ott: And we delved into that in a podcast episode we did on following the digital thread, where we actually spent some time following a digital thread through an airplane part. One little piece of it.
Male voice 1: We have been talking about this for decades.
Female Voice 1: We’ve only just begun to imagine the power of what this will unleash.
Male Voice 2: It compresses the supply chain from days, weeks, months, is some cases to zero.
Male Voice 3: It allows parts to be created like they’ve never been created before.
Duncan Stewart: The point I want to make that is incredibly important and ties in directly into your digital thread: If all of the manufacturing in the world moved to 3D printers tomorrow—which is not going to happen—that would be an enormously difficult challenge, but it would be a challenge of complexity X, let us say. If, on the other hand, we didn’t have any 3D printers and we made everything the old-fashioned way, that would be another kind of complexity. The reality is it’s much worse than that. People will be making prototypes on 3D printers. Then they will make maybe one or two parts on a 3D printer. Then they will using digital files make more parts on a CNC, a computer numeric control machine. Then when you want to make millions of parts, you do it on traditional manufacturing.
But as you get near the end of life or you have a request for a custom part, you might want to use a CNC machine or you might want to use a 3D printer and it will vary over time and by country. Controlling the digital thread among all of these different manufacturing technologies is an order of complexity that is X squared or X to the tenth or something like that. The most fascinating challenge that 3D printing brings to the market is how much more complex it makes this entire digital thread conversation. That’s a real opportunity but it’s also a challenge for many companies now.
Tanya Ott: Duncan Stewart, thank you so much for your time. Thanks for helping us understand this really tough stuff.
Duncan Stewart: Thank you very much.
Tanya Ott: Duncan Stewart is the director of Technology, Media and Telecommunications Research for Deloitte Canada. He co-authored this year’s TMT Predictions 2019 report, which you can find at deloitteinsights.com. You’ll also find that video series on the digital thread there, as well as our archives of this podcast—hours of conversations about some of the biggest issues facing business today.
Follow us on Twitter at @deloitteinsight. I’m on Twitter at @tanyaott1. Thanks for listening and have a fantastic day!
This podcast is provided by Deloitte and is intended to provide general information only. This podcast is not intended to constitute advice or services of any kind. For additional information about Deloitte, go to deloitte.com/about.
Today, business and technology innovation are inextricably linked and the demand for technology-enabled business transformation services is rapidly growing. Deloitte technology professionals around the world help clients resolve their most critical information and technology challenges.