Winfried Hensinger likes Star Trek. “It goes all the way back to primary school,” said the director of the Sussex Centre for Quantum Technologies in England. “I wanted to be science officer on the Enterprise, so I worked out in about grade five that I wanted to study physics.”
Today, his day-to-day work on abstract notions of quantum mechanics would make even Spock’s ears perk up.
“[Quantum computing] has a huge appeal for young people,” Hensinger told Digital Trends, “because it’s basically science fiction.” When he started in the field, it was largely confined to theoretical study. Today, the most promising projects are within reach of producing a universal quantum computer — something that was as sci-fi as Star Trek just a few years ago.
“Before there were computers, I had to learn typing on a typewriter,” Hensinger said with a laugh. “Life really changed when computers became available. And quantum computing could be a similar revolution.” When that revolution comes, it will be the result of several decades of work from committed scientists, mathematicians, and engineers. But what IS it, exactly? Ah, there’s the story.
We’re now seeing a race toward the first large-scale universal quantum computer.
Time Magazine described quantum computing beautifully in a 2014 cover story called “The Infinity Machine”: “It promises to solve some of humanity’s most complex problems. It’s backed by Amazon founder and CEO Jeff Bezos, NASA, and the CIA. Each one costs $10,000,000 and operates at 459 degrees below zero. And nobody knows how it actually works.”
As the quantum computer makes the jump from theory to reality, the field surrounding it is poised to explode. We’ll see applications that were never even considered before, from science lab fare that’s only comprehensible to experts, to weird and wonderful ways of using the tech that should capture anyone’s imagination.
And it’s all thanks to a very famous cat.
A Quantum of Science
Think about this: The device you’re using to read this article works with information stored as binary digits, or bits, each of which can occupy a discrete value of 1 or 0. A standard character from the Latin alphabet is made up of eight bits, otherwise known as a byte. The eight bits in a byte, each a 0 or a 1, can together refer to any alphanumeric symbol. Everything a computer processes, no matter how complex, can be reduced to a string of bits.
But a quantum computer works with information that’s stored in quantum bits, or qubits. A qubit can occupy a value of 1, 0, or any quantum superposition of the two states. That makes things a little more complicated.
Quantum superposition is an example of the “quantum weirdness” that scientists have grappled with for decades. Put simply, it means a quantum object can occupy more than one state until it’s measured, as referred to in the famous thought experiment of Schrödinger’s cat.
Instead of using transistors to keep track of binary data values, a quantum computer uses quantum objects. As a result, the computational power of a quantum computer easily eclipses that of a classical computer. Any given set of quantum values can represent far more data than traditional binary data, because it doesn’t have to reduce data to a string of 0s and 1s.
In theory, at least. Though researchers have been able to agree on how a quantum computer might function, building working hardware has proven an incredible challenge, and has led to plenty of disagreement.
Right now, we’re at a stage where researchers are able to construct a system that has access to a handful of qubits. These computers are great for testing out things like hardware configurations and even running algorithms, but as Time pointed out, they’re ungodly expensive and only the most basic version of what researchers envision. The true potential of a quantum system will only be realized once hundreds, or thousands, of qubits can coexist together.
We’re now seeing a race toward the first large-scale universal quantum computer. There are two strong but different contenders, and it’s not clear which idea will become reality first.
Theory into Practice
“It used to be a physics problem. Now, it’s an engineering problem,” said Hensinger, when asked about the changing face of quantum computing. Though there’s a clear theoretical understanding of how a quantum computer should work, that doesn’t make constructing one an easy process.
The difficulties are multifaceted, and researchers have yet to agree on what the foundation of quantum computing should look like. However, two possibilities stand head and shoulders above the rest.
One is called “superconducting qubits.” It uses an implementation that relies on supercooled electric circuits, and could offer manufacturing advantages when chips are being made in greater quantities.
The other is “trapped ions,” a method that is less confined by environmental factors like temperature, but faces other challenges, such as controlling numerous individual charged atoms inside a vacuum.
“Both superconducting qubits and trapped ions are very much an elite,” said Hensinger. “They’re both really good implementations, and they will both produce a quantum computer.” He and his team at the University of Sussex have opted to pursue trapped ions, a decision that was very carefully considered.
“I closely investigated all the different implementations and their progress,” Hensinger explained. “I tried to understand their prospects and their difficulties.” His investigation showed trapped ions were a nose ahead of the competition. Trapped ions are well-isolated from environmental noise that may destroy quantum effects, an advantage that Hensinger hopes will make the technology easier to implement, as well as easier for others to understand.
The method invented by Hensinger and his team at Sussex is based around individually controlled voltages that are applied to implement the quantum gates that make up a quantum circuit. This new technology constitutes a bold departure compared to how quantum gates with trapped ions have been implemented in the past — utilizing laser beams — tremendously simplifying the engineering required to construct a large-scale quantum computer. As such, the process remains in the realm of means-tested methodology ironed out over the course of decades of classical computing by imitating traditional transistor architecture rather than throwing out the rulebook completely.
“With this concept, it becomes much, much easier to build a quantum computer,” added Hensinger. “This is one of the reasons why I’m very optimistic about trapped ions.”
There’s another, even more practical reason the team at the Sussex Centre for Quantum Technologies chose trapped ions – it works at room temperature. Superconducting qubits doesn’t.
The Other Path
IBM is one of the largest companies competing in the emerging field of quantum computers, and its size provides it advantages. First and foremost, there’s the matter of funding, which is an easier proposition for a company worth billions than it is for an academic organization working within its means.
Money isn’t the only advantage that IBM enjoys over its competition, though. Over the course of the 105 years that the company has been in business, plenty of powerful minds have passed through its halls — and, in some cases, their expertise is still hard at work. The company’s combination of resources and talent has led it to choose superconducting qubits as the basis for its work.
The physical construction of IBM’s quantum computer is comprised of a five-qubit processor that’s housed in a printed circuit board. “That’s housed in the bottom of our dilution refrigerators,” said Jerry Chow, manager of the Experimental Quantum Computing team at IBM. “If you see pictures of them, they look gigantic, almost like beer kegs.”
Throughout the refrigerator, control wires carry microwave signals down into the chip, and route signals out through a series of amplifiers and passive microwave components. These messages can be interpreted with a classical computer to allow the team to read the system’s qubit states from outside the fridge.
“I think superconducting qubits are really attractive because they’re micro-fabricatable.”
Between the refrigerator and all the supporting electronics required, the set-up for IBM’s quantum computer takes up more than 100 square feet of laboratory space. That’s the downside to the technology. It requires incredible cooling, and that cooling isn’t easy to miniaturize.
But why go to all that trouble, when trapped ions provide an alternative?
“The reason that IBM is using superconducting qubits is kind of historical,” said Chow. “There was a program here that worked on low-temperature classical computing using Josephson junctions.” Interest continued among researchers who had the necessary expertise, and when ideas about superconducting circuits for quantum computing emerged, the company’s previous low-temperature work helped it hit the ground running.
And there’s more. IBM’s chosen technology aligns well with its expertise in building computer chips. “I think superconducting qubits are really attractive because they’re micro-fabricatable,” Chow explained. “You can make them on a computer chip, on a silicon wafer, pattern them with the standard lithography techniques using transistor processes, and in that sense have a pretty straightforward route toward scaling.”
These similarities don’t mean there’s similarity between IBM’s quantum computer and your work laptop — the technology is very different. But certain elements of the manufacturing process resemble techniques that are prevalent today.
IBM and the University of Sussex find themselves among the organizations racing toward the end goal of a large-scale universal quantum computer. But the individuals involved with the research aren’t looking for bragging rights. They want to make history.
For a scientist who has been working on quantum computing, the question of whether the finished product uses trapped ions or superconducting qubits is far from a priority.
“It’s like going back to 50 years ago, when computers were still filling whole buildings, and asking whether you want a Windows PC, or a Mac,” said Hensinger. “I just want a quantum computer.”
Trapped ions and superconducting qubits have emerged as the two frontrunners.
The field has witnessed numerous research groups crash and burn despite early promise. Now that trapped ions and superconducting qubits have emerged as the two frontrunners, it’s beneficial to all interested parties that the hopes of a large-scale quantum computer aren’t tied to the potency of one particular solution.
“It’s healthy to have different groups trying different things,” noted Hensinger. While there is a sense of accomplishment attached to being part of a group that’s made significant progress, all findings are shared with the wider scientific community, so the field as a whole can move forward.
At the moment, the team at the University of Sussex is combining several different experiments into one feature-complete system that has access to a handful of qubits. This project is expected to be complete in three to five years. Beyond that, a large-scale quantum computer that uses many qubits is the goal, though that could be 15 or 20 years away. IBM has a similar time frame for its project.
Both research groups are set to make more incremental advances in the interim. “Over these next few years, you’re going to see coherence times improve, and you’re going to see the number of fully controllable qubits increase as well,” Chow explained.
It’s impossible to know the full scope of what quantum computing will be used to accomplish, given that the ideal realization of the technology is still years in the future. But specialists all over the world are working with quantum algorithms to see what the hardware could be capable of, and the results could have as much of an impact as the hardware.
Krysta Svore was first exposed to quantum computing when she was studying math under Andrew Wiles. “He mentioned that there was a model of computation — a potential computer — that could actually break RSA encryption,” said Svore. “We call this the factoring problem.”
“I was a math major, so I was completely fascinated that there was this computer that could break this problem that we thought to be very hard,” she explained. Svore dove in headfirst, and today she manages the Quantum Architectures and Computation Group at Microsoft Research. Her team’s focus is software, but that spans from the development of specific quantum algorithms to work on infrastructure like software architecture and toolsets.
But what is a quantum algorithm, and what does it do?
“A quantum algorithm is going to be, at the highest level, similar to a classical algorithm, in that it’s a recipe for solving a problem,” explained Svore. “And that recipe is often written in terms of mathematics.”
While a human brain can understand a query like “find all the whole number factors of X,” a computer won’t. Very specific binary input is required to make a computer solve even the simplest equation. Coding in binary is inefficient, however – so programmers working with classical computers use computer languages like Python or C to span the gap.
Coding languages have become extremely advanced, allowing most programmers to work without touching raw binary inputs and outputs even once. Entire algorithms are often coded into programming languages, so they can be used simply by inputting the proper reference.
Yet all popular programming languages are built to translate instructions to binary, machine code. Since that code isn’t what’s used by a quantum computer, they’re useless for programming a quantum computer. It’s the job of Svore, and her team, to come up with new languages that will work.
It’s imperative that the quantum software is given attention now, so that the field doesn’t hit a wall once hardware challenges are under control.
“[We have a] programming language that we have developed explicitly for quantum computing,” said Svore. “Our language, and our tools are called LIQUI|>. LIQUI|> allows us to express these quantum algorithms, then it goes through a set of optimizations, compilations, and basically rewrites the language instructions into device-specific instructions.”
Microsoft has its own platform in place, but a similar process is being used by research groups around the world. Dorit Aharonov is a professor at the Hebrew University of Jerusalem, and her description of an algorithm running on quantum hardware echoes Svore’s comparison to a recipe.
“An algorithm is a sequence of small steps which the computer performs, and for the algorithm to be efficient, it needs to perform the task with as few as possible such steps,” she explained. “The whole point is that quantum algorithms manage to perform certain computations much faster than any classical algorithm, because of their ability to explore exponentially many possibilities at once.”
So, if an algorithm is a set of steps used to carry out an operation, a quantum algorithm is a similar sequence of instructions specifically designed to harness the computational power of quantum objects. However, there’s another complication to the relationship between software and hardware — without large-scale universal quantum computers out in the wild, it’s difficult for researchers working on algorithms to ensure that their work is headed in the right direction.
It’s imperative that the quantum software is given attention now, so that the field doesn’t hit a wall once hardware challenges are under control. In many cases, hardware and software research projects are working in sync.
Coding for Quantum
In broad terms, quantum algorithms are the software to the hardware of a quantum computer — but this analogy only goes so far. Given that the hardware side of things is still in such a state of flux, the development of quantum algorithms is informing the ongoing development of the quantum computer itself, and vice versa.
Microsoft Research works with world-class teams at the University of Copenhagen, Delft University of Technology in the Netherlands, and the University of Sydney, so it’s hardware is located in academic labs around the world. “We visit them regularly,” Svore said. “We’re pretty much a virtual team at the moment, because some of the best resources are global.”
The IBM Experience is designed to open up the hardware to everyone.
As well as physical hardware, the Quantum Architectures and Computation Group has access to a simulated quantum computer that’s running on a classical system. “This allows us to debug quantum algorithms, to test and design new quantum algorithms, new quantum circuits, and subroutines,” said Svore. “We want to test the device design and algorithm design as much as possible in simulation, before running it on the actual device.”
This simulation offers enormous advantages to the team. Svore and her associates can test out their algorithms and iterate as necessary, which in turn allows the hardware teams to see how quantum computers will be used in practice. But, as you might imagine, simulating quantum hardware on a classical computer has its limitations. The virtual device that Svore and her team use simulates around 30 qubits for 32 GB of RAM, and every time an extra qubit is added for simulation, that memory has to be doubled.
Microsoft isn’t the only major company that’s using applied usage to means-test its quantum output. Earlier this year saw the launch of the IBM Experience, an online interface that allows academics and enthusiasts to take the reins of a five-qubit system.
“We wanted to […] build a community to further develop research in this field,” said IBM’s Chow. “We thought that, as we start making these quantum processors bigger and with more functionality, at some point it’s not going to be something that everybody has the ability to start up in a new lab. We wanted to show the quality of these processors that we’re building and the system that we have around them, but also to enable research on them and accelerate innovation in the field of quantum computing research.”According to Chow, users have already used the IBM Experience to program in basic quantum mechanics tests, simple error detection protocols, and experiments observing noise levels compared to a simulation. The program offers unprecedented access to working hardware, and that’s a huge boon to students hoping to enter the field.
“From an education standpoint, this is a really big thing,” Chow added. “We’re able to give access to students who are in the midst of a quantum computing course, or a quantum mechanics course, and they’re able to tinker with it and kind of reaffirm what they’re learning in their class but in a live system.” Knowing the tool would serve users of all experience levels, the team made an effort to make it as welcoming as possible.
“We wanted to have a very visually appealing, visually striking feel to it,” Chow explained. “That’s why we made a Composer tool — we wanted to have a strong analogy with music.” The tool is used to create quantum score files, which look like a musical register. There are five lines that respond to the five qubits of the computer, and users can drag and drop different gates and operations from the library onto each.
“You’re able to visually see what a quantum program looks like before you run it,” said Chow. The comparison between a musical performance and a quantum computer is rather prevalent in the field, so it’s only appropriate that it would be called upon to ease non-experts into the process. “It is a good analogy in the sense that time matters, and so does the order of operations. So, in that sense, it really is like a musical score.”
Don’t go feeling left out if you feel like you would never be able to get your head around using a quantum computer — only a miniscule percentage of the world’s population has been given the reins to one as of 2016. Most of the people who have used such a system to date have probably been rather intimidated by it.That’s something that something engineers like Svore want to change. The LIQUI|> platform is designed to give people in the field right now a relatively straightforward method of interacting with quantum hardware. The IBM Experience is designed to open up the hardware to everyone, whether they’re familiar with the underlying concepts or they just have an interest in the tech.
While hardware development continues toward the end goal of a large-scale universal system, these efforts help to bring even more talented minds into the field. These early software efforts hope to build an audience of interested and skilled coders who can jump right into quantum computing.
What sort of applications will they be working on? The possibilities are endless. But to get an idea of where the field is at — and where it could be headed — it’s worth looking back at the algorithm that jump-started the interest in quantum computing.
“Shor’s Algorithm is definitely the most famous and spectacular application of quantum computing,” said Ryan O’Donnell of the Carnegie Mellon School of Computer Science in Pittsburgh. “Really, it was the major inspiration for the field.”
As long as there have been computers — whether human or electronic — they have been called upon to find the prime factors of a given number. This isn’t too difficult when the numbers are small and manageable, but the higher the value, the more complex the process becomes.
“If a full-scale quantum computer could be built, it would mean that that computer could break almost all the cryptography used out there.”
“I’m talking about superenormous numbers, numbers with thousands of digits,” said O’Donnell. “It’s extremely easy for computers to multiply two 10,000-digit numbers. It takes nanoseconds. But we don’t know any efficient algorithm for computers to do the opposite problem — taking a number and finding its factors.”
Efficient is the key word. There is an algorithm that’s capable of factoring huge numbers, but unfortunately it’s nowhere near efficient enough to do the job. In the time that it would take for it to factor a 10,000-digit number, the life span of the universe could play out in its entirety. A computer that’s 10 times faster, or even a thousand times faster, offers no benefit. A completely new form of computing is required.
“In 1997, Shor showed an algorithm for a quantum computer that would be able to factor such numbers very efficiently,” said O’Donnell. Since then, it’s become something of a measuring stick for the advancement of the field as a whole.
In 2001, IBM managed to turn the number 15 into its factors, 3 and 5, using a seven-qubit quantum computer. By 2012, a research team had managed to factorize the number 21. Then, in 2014, a more advanced method known as the minimization algorithm successfully factorized the number 56,153.
“Besides the practical implications, Shor’s algorithm is also just very mathematically beautiful!” O’Donnell added. “It inspired a lot of people to look for superfast quantum algorithms for other computational tasks we can’t solve efficiently with normal computers.”
As you can imagine, looking for those tasks requires a certain amount of abstract thinking, as well as a strong grasp of the complex concepts involved. Coming up with applications for a form of computing that’s still under active development is no small feat.
“One might wonder about the value of doing this before quantum computers are even built, but I’m reminded of the story of Ada Lovelace and Babbage’s ‘Analytical Engine,'” said O’Donnell. Hypothetical plans for a general-purpose physical computer were in place as early as the 1850s. “Ada Lovelace famously designed algorithms for solving certain math problems on Babbage’s computer, 100 years before it could even be built.”
It can be difficult to see past the abstract nature of a quantum algorithm, and understand how the concepts could be applied to solve real-world problems — especially if you don’t hold a degree in mathematics. Fortunately, Shor’s algorithm can demonstrate the results, or perhaps consequences, of the quantum computer.
Consider Shor’s algorithm running on a classical computer. As mentioned previously, the factorization of a 10,000-digit number would take many millennia to play out. “Although that sounds unfortunate, in one way, it’s actually very fortunate,” said O’Donnell. “Almost all internet security and cryptography relies on the fact that no computer can factor 10,000-digit numbers in a realistic amount of time.”
It’s not too difficult to guess what quantum computers will mean for cryptography if the hardware becomes viable. “If a full-scale quantum computer could be built, it would mean that that computer could break almost all the cryptography used out there.” O’Donnell added.
Of course, a full-scale quantum computer is currently beyond the grasp of the leading research groups in the field, so a hacker collective isn’t going to build one for unsavory purposes just yet. But this should illustrate how the realization of quantum computing will have far-reaching ramifications. “Things are still secure, for now,” said O’Donnell. “But the National Security Agency has already suggested that everyone should move away from factoring-based cryptography in the near future, precisely because of the possibility of quantum computers being built.”
“Most people are looking at post-quantum cryptography from the perspective of, ‘how do we go about finding a classical encryption technique that is robust against a quantum attack,'” said Svore. “That’s the focus right now, to find something post RSA. What’s the next generation? We know that RSA is not robust to quantum computers — what are the others?”
It’s capable of fully simulating any arbitrary system. That’s something tremendously important.”
Another immediate application for quantum hardware is to further research other areas of science. “It’s capable of fully simulating any arbitrary system,” said Winfried Hensinger. “That’s something tremendously important.”
The limited computational power of a classical system means that certain elements of a simulated model have to be less precise than they would be optimally. “The difference with a quantum computer is that you don’t have to make those approximations anymore,” Hensinger continued. “You can just simulate the whole system.”
This could offer up some profound advantages for research projects that rely on simulation to gather data. Indeed, we’re already seeing meaningful progress toward this kind of usage. In July, Google announced that its quantum device was successfully used to calculate the electronic structure of hydrogen molecules.
The Quantum PC
“When I first entered the field, if I mentioned quantum computers, I got a very perplexed look back,” said Svore. “In other words, not many people had heard about quantum computers. Now, there’s cover stories on the New York Times, a quantum computer was on the cover of Time magazine.”
Quantum computing will have a palpable impact on our lives, even if it’s a relatively subtle change to our day-to-day affairs, like a new and improved form of encryption. As for a “quantum PC” sitting next to the inkjet printer in your office, that looks less likely, at least in the short term.
“Not everyone is going to own a quantum computer,” said Hensinger. “This is not the sort of system that you’re going to do word processing on.”
But that’s not to say the technology won’t, in some form, end up in the average household. The lineage of computer technology that brought about the ubiquitous laptops, desktops, and hybrids of today didn’t seek to create devices that could surf the web or play video games. These systems were designed to solve mathematical problems, things like the behavior of sea tides and the trajectory of ballistic weaponry — but their power was ultimately harnessed for more mundane tasks.
Imagine a new version of Rollercoaster Tycoon, where the likes and dislikes of every kid in the theme park are determined by a fully simulated childhood. Imagine a future version of Photoshop with brushes that behave with all the quirks of a real, physical implement. Imagine a version of Google Maps that can precisely calculate the time of your arrival, down to the second, based on a simulation of historical traffic patterns. Imagine just about anything — because someone will have to imagine these applications before they’re implemented.
Hensinger compared its potential to the jump from the typewriter to a PC.
Over the past decade, the quantum computer has transformed from a potent concept into a workable end goal. The next 10 years are going to be about reaching that end goal. While that process continues, new faces entering the field and plenty of those already working within it are going to have ideas about the ways that this tech can benefit everyday life.
You won’t need to install a dilution refrigerator in your basement as part of your next PC build. However, given the advances that are set to be made in the not-so-distant future, the ideas at the core of today’s quantum computers may well be integrated into consumer hardware.
Perhaps it won’t happen in the next decade, or even the next one after that — but the work being done over the next 10 years will be instrumental to the overall process, even if its effects aren’t immediately obvious to the general public. The promise of quantum computing extends far beyond the hardware that’s currently viable. That’s why Hensinger compared its potential to the jump from the typewriter to a PC. And he’s not alone in his optimism.
“It’s going to really revolutionize computing,” said Svore. “And we cannot pass up this opportunity.”
- Chess. Jeopardy. Go. Why do we use games as a benchmark for A.I.?
- Evolving, self-replicating robots are here — but don’t worry about an uprising
- Image-recognition A.I. has a big weakness. This could be the solution
- Silicon Valley is racing to build an audio-only internet, and I hope it succeeds
- How to recycle your old computer