Qutech: building a quantum delta, one (qu)bit at a time
Even though quantum technologies are only just starting to take shape, now is the time to consider the opportunities they present - both to enhance existing business and to generate new opportunities. In the Netherlands, the Qutech initiative is leading the effort to bootstrap a quantum industry in the Dutch delta.
Author: Paul van Gerven, Bits&Chips
The quantum computer, despite all the excitement surrounding it, is still a mysterious thing. Sure, the theory is solid: scientists have a pretty good idea of what’s needed to build one. But those are the contours, the basic principles. Few would dare to predict what the first full-fledged quantum computer would actually look like on the inside, what technology it would be based on. Even today, several candidates are being considered, and probably more options will pop up – or fizzle out - as time passes.
Still, the quantum computer has moved from concept into reality: already some primitive ones have been built. To outsiders, these systems can’t do anything particularly useful, but the point is that quantum calculations are being performed, which means people are getting their hands dirty.
These quantum builders need to think about the whole system: not just the quantum bits (qubits) that perform the calculations but also the electronics and software that are essential to manage and control them. Just like a PC needs a hard drive and an operating system in addition to a processor, a quantum computer needs peripheral electronics. Actually, it needs a lot of it. It takes a couple of pretty sizable ‘classic’ computers to run a quantum one.
The elements that constitute the modern computer were developed gradually, one at a time, until they all came together. It started decades ago with the invention of the transistor, then came the integrated circuit (computer chip), at some point, someone started working on the software, and so on.
That’s not how the quantum computer is being built, explains Ronald Hanson, scientific director of the Dutch quantum research center Qutech in Delft. “Already at this early stage, we have many disciplines working together, both scientists and engineers. Making progress in one particular area depends on progress being made in all others as well. It’s not a step-by-step evolution of isolated technologies.”
Qutech is one of the few places in the world where such a concerted effort is taking place. Google and IBM are the most well-known organizations building complete quantum computers, along with some lesser-known companies, while most research groups focus on a single problem or a few at most. But Qutech is neither a company nor a research group. It’s a public-private collaboration trying to leverage its efforts into establishing a quantum industry in the Dutch delta. How has that been working out, so far?
What makes a quantum computer so great?
In a regular computer, information is coded into a series of 1s and 0s, the bits. These, in turn, correspond to physical states. The 1s and 0s on a hard disk, for example, correspond to the direction of a magnetic field (up or down) on a large number of distinct areas on a surface. A computer essentially performs operations (manipulations) on a series of 1s and 0s according to the instructions given by the user.
A quantum computer, too, uses physical states to represent 1s and 0s, but – crucially - these physical states are small enough to obey the laws of quantum mechanics. It’s best for anyone without a degree in physics not to try and understand quantum mechanics, so you just need to know that quantum mechanics allows physical systems to be in two states at once. So, in a quantum computer, where a bit is called a qubit, there’s the 1s, the 0s and the both-1-and-0s.
The both-1-and-0s are where the magic comes from: you can perform two operations for the price of one, for example a single operation can be performed on both a 1 and a 0 at once. A normal computer needs two operations for that.
One or two operations – it may not seem like a big deal. But the difference between quantum and classical computing grows exponentially. A 10 qubit quantum computer can already perform 210 = 1024 operations at once, so imagine the power of a quantum computer with hundreds of qubits!
There is a caveat, however: qubits are extremely fragile and frequently ‘flip’ their value. This means extensive error correction is required, which is made possible by using many physical qubits to make a single ‘real’ qubit. Currently, researchers expect the ultimate quantum computer will need millions of physical qubits, whereas the biggest one now only has tens of (still very fragile) qubits.
Another important thing to note is that not all types of calculations benefit from quantum power. A quantum computer will be great at finding out that 268,568 is 569 times 472, for example. A classical computer has to check every reasonable combination of numbers until the solution is found. Not every problem is necessarily of this nature, though.
Still, there are plenty of applications for quantum computers. They could be used for machine learning, in financial modeling and for logistics, for example. Another great application for quantum computers is in chemistry. Molecules are quantum systems and chemical reactions are dictated by quantum mechanics. What could better simulate these systems than an actual quantum computer? Imagine being able to design new drugs or materials without ever setting foot in a lab.
Qutech was founded in 2013 by TU Delft, a university of technology and engineering, and TNO, an applied research institute. TU Delft had world-class quantum research to offer, as well as expertise in electrical engineering, physics and computer sciences – all very relevant for quantum technologies. TNO knows a thing or two about those things too, having assisted the international high tech industry in innovation for decades. Another role of Qutech, however, is to reach out to companies and try to get them involved with quantum technology.
Two big fish already signed up: Microsoft, which built its own lab next to Qutech’s premises, and chip manufacturer Intel, which entered into a long-term partnership. Of course, these are the kind of companies that don’t need convincing that quantum tech is a paradigm-shifting opportunity. “Getting firms on board whose core business doesn’t involve computing as much is a lot harder,” admits Rogier Verberk, director Semiconductor Equipment Industry at TNO and the organization’s point man for Qutech. “But we’re getting more and more traction.”
Qutech is one of many research initiatives co-funded by the Top Consortium for Knowledge and Innovation (TKI) High Tech Systems and Materials. Through the Ministry of Economic Affairs and Climate, TKIs provide additional funding whenever industry invests in long-term innovation projects with public research organizations.
Roughly three types of companies are potential Qutech partners. First, perhaps surprisingly, there are the end users – the companies that could take advantage of quantum technology. That may sound strange, given the primitive state the technology is still in, but that shouldn’t deter anyone, Hanson assures. “The perfect quantum computer may be years or even decades away, but before we get there, we may be able to solve real-world problems with imperfect quantum computers.”
Hanson refers to what are called noisy intermediate-scale quantum (NISQ) machines. These are modestly sized systems in terms of computing power and furthermore prone to making errors. Even if lacking compared to the real thing, NISQs could still prove useful for a range of problems that are out of reach of regular (super)computers, because methods can be devised to compensate for the errors.
“The problem is that while we know what those problems look like, we don’t know many that would actually be useful to solve. In fact, it has become a research field on its own to find those,” says Hanson. “Surely, though, many such problems are to be found at companies. At banks perhaps, to sift through data. Or at pharmaceutical companies, looking to design a new drug. These companies have no idea what quantum computing could do for them, however.” Which means there’s all the more reason for Qutech to actively reach out and get companies involved.
Already Qutech joined forces with a bank, ABN AMRO, though not for quantum computing but for quantum communication, a branch of quantum technology that’s a little farther along than computing. The project involves exchanging extremely secure cryptographic keys between users in a way that’s practically impossible to eavesdrop on. In the future, these keys could be used to secure Internet and mobile banking. Taking this one step further, Qutech is also working with telecom company KPN to make the quantum Internet a reality.
Quantum Delta NL
Qutech may be the front-runner, but many more Dutch organizations are involved in building the Quantum Delta NL. Most universities have some kind of quantum research program, Qusoft from Amsterdam focuses on application software for quantum computers and the quantum Internet, and the collective of Dutch high tech companies called the top sector High Tech Systems and Materials (HTSM) is keen to stay involved. In September 2019, these organizations presented their joint goals and ambitions in the National Agenda Quantum Technologies. Apart from the commercial ecosystem, the agenda also addresses educational and human capital issues as well as informing and engaging the general public.
The second kind of company that fits in well in the nascent Dutch quantum delta is one that’s involved in the quantum business itself. Qutech is the reason Finnish company Bluefors decided to open an R&D office on the TU Delft campus, where Qutech resides. Bluefors designs cryogenic systems for quantum computers. “It’s important for us to be able to design new specifications with leading users and to benefit from each other’s knowledge,” commented Bluefors CEO Rob Blaauwgeers when the Delft office was announced in 2018. More companies might want to follow this example.
Qutech itself gave birth to two quantum startups already, both of which are now operational in Delft. The Qblox spinoff focuses on the previously mentioned specialized electronics and software required to run a quantum computer. Hanson: “You can’t operate a quantum computer using a normal PC. At Qutech, we’ve put a lot of work into electronics that satisfies the specific demands for quantum computing. Our results are so good that it makes sense to start selling.” The other spinoff, Delft Circuits, focuses on dedicated electronics wiring for quantum computing at ultra-low temperatures. A third spinoff is in the making, but Hanson can’t tell anything about that one yet.
The third and final type of ‘Qutech company’ may be considered both a potential user and a technology supplier. Many high tech companies will be able to take advantage of quantum computing once it reaches a certain performance level, enhancing their toolkit for innovation. At the same time, quantum computing will present new business opportunities. As the technology evolves, there’s likely more revenue to claim. The supply chain will increasingly become more sophisticated, just like that of – say – the PC is very large and complex these days. “We’re basically shooting for something very similar to the semicon ecosystem we already have here in the Netherlands, with companies like ASML, NXP and their supply chains. And we want to be the best in Europe,” explains Verberk.
Wait, aren’t you the best in Europe already?
Verberk, smiling: “We don’t like to be immodest, but yes, there’s something to be said for that. The large number of people that work here, the world-class quality of the research, the fact that we successfully brought together scientists and engineers under one roof, and that Microsoft and Intel chose to join ... let’s just say many people admire what we’ve done here.”
Still, there are quite a few places in Europe that aspire to become a quantum hub. That’s a lot of competition.
Hanson: “Ultimately, a quantum hub needs critical mass to succeed. Clearly, we can’t have twenty hubs all over Europe. I suspect we’ll see consolidation over the next few years. In fact, I hope the European Union will encourage this. And I’m confident that Qutech will be one of the ecosystems to make the cut.”
Are companies already lining up to start working with you?
Hanson: “Compared to other countries, such as Germany and the United States, I must say that Dutch companies have been quite cautious. I’d really like to see things speed up. We need companies to get on board and, reversely, I don’t think it’s an exaggeration to state that this technology is too important to miss out on.”
“On the other hand, we have some improvement recently. Before we started Qutech, I invited twenty companies or so, gave them a tour and told them what we were setting up. Afterward, I asked them whether they supported our concept, whether they would like to stay involved and whether they would like to invest in quantum technology. They all liked our idea and wanted to stay involved, but that was it. Nowadays, their interest is much more tangible. Only a few years ago, ABN AMRO would never have invested in quantum technology. And recently, we signed a long-term collaboration agreement with KPN.”
Verberk nods: “Not that long ago, quantum technology wasn’t an item at trade conferences, not even at technology conferences. Today, dedicated parallel sessions are popping up all over the place. Slowly but surely the awareness grows. Our meetings with companies used to be generally exploratory in nature. These days, we dig a lot deeper.”
It must be hard to convince companies without having something to show them, though.
Hanson: “We have demonstrator programs! We’re building a small quantum Internet and we have a quantum computer that outsiders can use for getting to know the technology and its possibilities. Right now, it’s a simulation running on a supercomputer, but we want to start offering real quantum hardware in 2020. Having such goals is an excellent way to drive our own efforts, by the way. Because everything will need to be on point. Every single part of the computer will need to be up and running and working well with other parts.”
IBM has had a similar program for a long time, on real hardware. Google recently had its quantum computer outperform a supercomputer on a specific problem for the first time in history. It seems like you’re falling behind.
Hanson: “What Google did, is a breakthrough. No doubt about that. Yet, I don’t think we can compare what we are and what Google is trying to do. Google chose a very specific challenge five years ago: to demonstrate quantum supremacy. They succeeded, which is fantastic. But that result isn’t very relevant to Qutech’s mission.”
“As for IBM, it did very well to start offering quantum computing to the public. That’s necessary for educational purposes and for community building – two things that are really important for a radical new technology. That’s very much like Qutech’s mission: educating people – TU Delft started a new series of master courses Quantum Technology – and building an ecosystem.”
Verberk: “We’re not competing with Google and IBM, and it’s not our mission to outdo Google and IBM. What matters is what we do to bring quantum technology to fruition in the Netherlands.”
There’s a long road ahead of you. What are the biggest obstacles?
Hanson: “Apart from getting companies involved, the biggest bottleneck is talent. Quantum technology needs a lot of qualified people, but there simply aren’t that many of them around. There’s a reasonably steady supply of young researchers, but experienced talent is hard to come by. That’s definitely going to be our biggest problem in the coming years.”