Breaker of Chains: The Appeals and Perils of Quantum Computing

In mid-December 2017, days after New York experienced its first snowfall of the season and temperatures dropped to freezing, IBM announced the launch of the IBM Q Network, bringing together a group of Fortune 500 companies, academic institutions and national research labs to experiment with potential real-world applications on its 20 quantum bit (qubit) processor. 

Included in this club are two heavy-hitting banks—JPMorgan Chase and Barclays. JPMorgan says it will use the network to experiment on use-cases for quantum computing applicable to the financial services industry, including trading strategies, portfolio optimization, asset pricing and risk analysis. Robert Stolte, managing director of JPMorgan’s Corporate and Investment Bank (CIB), says that they are looking to be an active participant as they explore how quantum computing will potentially be applicable to CIB’s business.

“Partnering with IBM Research in the Q Network is a great opportunity for us to bring our engineers alongside IBMs research team,” Stolte tells Waters. “Together we can leverage IBM’s existing quantum computers, and those that are being developed, to explore the types of problems that this technology may be able to help solve in the future.”

He adds, though, that this is just a starting point—immediate results should not be expected.” This partnership is about the future and learning together how to think about the potential for quantum computing in our industry; it isn’t about short-term application,” he says.

Barclays, on the other hand, will build its knowledge of general approaches to quantum computing and start investigating potential use-cases in finance, according to Dr Lee Braine, who works in the investment bank CTO office at Barclays, which will lead this project at the firm. “We are keen to investigate the latest advances in quantum computing and how these leading-edge technologies could one day add value to Barclays and its clients,” Braine tells Waters. “We will begin with experimentation. We will take some of our existing optimization challenges and using the IBM Q Network, experiment on how quantum computing could potentially help.” 

Major Step Forward

To have two global banks like these throw their respective hats into the quantum computing ring marks a major step forward, but they’re by no means the first. Sophisticated quant hedge funds Two Sigma, Renaissance, DE Shaw and WorldQuant have begun to test these tools, according to the Financial Times. And in April 2017, the Commonwealth Bank of Australia (CBA) announced that it would begin experimenting with quantum technology by partnering with specialist firm QxBranch, which has a quantum simulator. (UBS is also working with QxBranch, but declined to comment for this story.)

Then, in August, CBA poured in some A$14 million ($10.7 million) into Australia’s first quantum computing company, Silicon Quantum Computing (SQC). Other founders include telecommunications firm Telstra, the Australian Federal Government, the New South Wales Government and the University of New South Wales. The aim of SQC is to develop and commercialize the first silicon-based quantum computer. 

While the likes of IBM, Google and Microsoft, as well as a few well-funded startups, have made impressive leaps when it comes to quantum hardware, the QC industry is still in its infancy and far from broad scale commercial adoption. 

Google claimed that it would have “quantum supremacy” by the end of 2017, with a 49- or 50-qubit machine. But IBM got there first, announcing in November that it had built and measured the first 50-qubit prototype processor, with a coherence time—the amount of time available to perform quantum computations—of 90 microseconds. 

It would appear that the financial services sector is now gearing up for what will be a revolutionary technology that will provide a lot more room for computations to be done simultaneously rather than in sequence or in parallel, as is currently the case on classic computers—the computers of today that we all know and love. 

While there is great potential for improved efficiency and the ability to solve previously unsolvable problems, the reason why these banks are getting involved so early on is because there are also existential questions that need to be answered relating to security and encryption. 

The Blockchain Question

A feature than ran in the August 2017 issue of Waters touched on the potential dangers that the rise of quantum computing might bring, the most prominent of which is Shor’s algorithm, which finds the prime factors of large numbers in an exponentially faster way than any existing algorithm. If unlocked, it would mean all current security measures for credit cards, emails and protected personal data would be thrown out the window. 

But what about the implications for something that is well-encrypted, such as distributed ledgers like blockchain? A CTO of a New York-based investment bank—whose firm is experimenting with blockchain and, as such, asked for anonymity—believes that the advent of quantum computing will render blockchain technology useless. “Qubits would blow away blockchain technology. While it’s still in its infancy, Microsoft just released a new language for using qubits. So this is speeding up a bit even if the hardware is still not there,” the CTO says.

Anthony Scriffignano, chief data scientist at consultancy Dun & Bradstreet, expresses concern about the advancement of quantum computers and what it will mean for encryption. “As you increase the number of qubits, there is more concern with stability because the qubits can influence each other through quantum entanglement,” he says. 

Quantum entanglement is when qubits interact with each other in such a way that when one is measured it will give an indication of the state of the other qubit, even when there is a large distance between them. “We need to seriously rethink encryption and seriously rethink things that rely on encryption, like blockchain,” he says. “Anything that is heavily dependent on complexity needs to be reconsidered in a quantum world.”

The reason encryption is so strong today is its complexity. It could take millennia to develop the key to read an encrypted message. It’s not that you can’t do it, Scriffignano says, it’s more a question of why would you want to do it when it would take far too much time and wouldn’t be worth the effort. 

Classical computers operate by simulating possible solutions to a problem linearly. Quantum computers would have the ability to try all of those keys at essentially the same time. “The complexity and the time it takes to decrypt a key doesn’t really go to zero, but it gets to a number that’s much smaller. This technique is called quantum hacking,” says Scriffignano.

However, Michael Brett, CEO of QxBranch, says he does not believe that the advancement of quantum computing will render blockchain technology useless. Rather, he says they will be “very complementary” technologies. 

“It will take a very advanced quantum computer, something on a scale we won’t see for at least a decade, to come close to rapidly unlocking blockchain-type cryptographic problems,” he says. “By then, blockchain technologies will have matured to be quantum resistant or incorporate quantum algorithms into their problems.” 

But, to Sciffignano’s point—and the reason why the likes of Barclays and JPMorgan are getting involved now—it is important to begin to understand how QC will affect technologies like blockchain, because you don’t want to make an investment today that will be rendered obsolete within five years. 

Keith Bear, vice president of financial markets at IBM, says quantum hacking is certainly perceived as a risk to blockchain encryption and raises the question of quantum proofing. 

“Blockchain technology is something that is talked about, but at one level, in the current state of quantum computing technology, that makes it somewhat of an academic point at the moment, but these are real concerns,” he says. “As we go into 50-qubits and beyond, that becomes much more significant.” 

Questions of Security

That said, quantum computing could also be used as a means of increasing security, Bear adds, as it can be used to manage the cryptography keys such that any “eavesdropping” that occurs can be detected.  “Obviously, it has a very high level of security because it assumes you observe a quantum state; it automatically takes one position or another,” he says. Here, Bear refers to the “No Information Without Disturbance Principle” where in quantum physics, if you measure a system, you change its state.

Giulio Chiribella, professor of computer science at the University of Oxford, and CIFAR-Azrieli global scholar, says this is one of the most important laws of quantum physics. “Every attempt to read the qubits in the intermediate steps would ruin the computation, making the quantum computer equivalent to a classical computer,” he says. “Between the beginning and the end of the computation, the quantum computer explores the exponentially large space of quantum superpositions.” 

From the Commonwealth Bank of Australia’s perspective, it has been looking at the complexity of Shor’s algorithm to see how it can be implemented on the simulator. It is also exploring post-quantum cryptography, cryptographic algorithms that are thought to be secure against quantum hacking, Dilan Rajasingham, head of emerging technologies at CBA, tells Waters. “If we assume tech is always increasing and we assume things are going to get easier and easier, how do we prepare ourselves for the future?” he asks. “What’s the algorithm that’s beyond RSA that’s quantum resistant? We’re doing this to make sure we can continue to guarantee the security and privacy of our customers and partners.” (RSA, which stands for Rivest-Shamir-Adleman, is one of the public-key cryptosystems and is used for secure data transmission.)

That said, at this point, he thinks the technology is still nascent, although CBA is trying to establish exactly where these tools are at through experimentation. “It’s too early to verify one way or another. If you understand the complexity of cryptography there are so many different cryptographic routines out there,” he says.

Working in Parallel 

Although, theoretically, quantum computing could “break” blockchains, what can be done to counter that in the future is to experiment with things like quantum key distribution, according to IBM’s Bear, as IBM is also actively involved in blockchain’s development, having launched a global banking payment platform in October.

“Quantum itself can be a means of increasing security in that respect,” he says. “In our point of view, it’s a key part of what we’re doing to understand and innovate the technology in the first place and to make sure that the existing cryptography and security that we’re implementing as part of the blockchain projects we’re working with are currently in the highest level of protecting security as you would expect. And obviously we’re going to be in a good position to affect that as a result of the work we’re doing from a quantum point of view as well,” says Bear.

“Obviously the security [of blockchain] itself will need to improve and enhance in order to cater for that risk as well,” he continues. “We have a direct incentive to facilitate that because we’re looking into the investment we’re making into blockchain. That’s where we expect this to develop in terms of maintaining and improving the levels of security within blockchain technology regardless of the risk and impact of new technologies like quantum computing.”

Quantum Machine Leaning

While security and encryption are most definitely areas to monitor and experiment with, there are other aspects of financial markets that banks such as CBA are working on within a quantum environment. CBA’s Rajasingham says the bank has a three-part strategy when it comes to quantum computing. The first is software, in which it is training its developers to become “quantum ready,” which is done through the use of the QxBranch simulator in experimenting with quantum applications. The second is hardware, and that comes with CBA’s direct involvement with SQC. This silicon-based architecture is one of the few paths to build a quantum computer. And the third is that it wants to experiment “broadly and deeply” with like-minded partners. When QxBranch deployed the quantum simulator at CBA’s premises, the invitation was left open to academics and others to experiment with the platform.

Rajasingham says that since the deployment of the simulator, there have been four main experiments they’ve overseen. “We’ve looked at applications for machine learning around big data analytics, quantum supremacy, blockchain consensus, and algorithms, which have been established for a while, such as Shor’s algorithm,” he says. 

However, the one that the CBA is immediately proceeding with is machine learning, as it is building its first application in financial portfolio optimization. “We’re looking at things like asset allocation,” he says. 

Although Rajasingham does not reveal a specific timeframe for when the experiment would be fully operational, he says that “at the moment, we would expect it to be ready in months,” hopefully by mid-year 2018. 

JPMorgan’s Stolte adds, though, that it’s still too soon to really know how quantum computing will interact with algorithms, in general, much less artificially-intelligent algos.

“Broadly speaking, there are many aspects of our business where we apply algorithms and financial models in computationally intensive processes that could be promising in the future,” he says.

Different Roads Travelled

But before the questions surrounding machine learning and blockchains can be answered, developers are still trying to figure out how best to build a quantum computer. There are a few routes one can take to build a quantum computer. These include superconducting electromagnetic loops, ion traps, photons, and single atom qubits in silicon, to name a few methods.

The likes of Google, IBM and D-Wave, a Canadian-based quantum computing systems and software company, are all using solid state systems superconducting qubits.

Chiribella says solid-state systems are “very promising” for their modularity. Other systems, like ion traps and nitrogen-vacancy centers, allow for higher fidelities in the  execution of quantum gates and longer storage of quantum information, respectively, he says.

This year could be the year that Microsoft unveils its Majorana qubit. Microsoft is working to build the first topological qubit, which it believes will be less error-prone and more suited for commercial use. This involves it manipulating Majorana fermions, which are subatomic particles.

It has also has unveiled a new programming language—Q#, which was mentioned earlier by the bank CTO—along with a toolkit to help coders create software for quantum computers.

QxBranch’s Brett says that while it’s very futuristic sounding, the science behind these architectures at a small-scale is extremely well established and the challenge today is in engineering bigger and bigger systems.

As for whether or not there is one type of architecture that best suits applications for financial firms, he says it’s still very early days. “That’s why companies like QxBranch are working with both financial institutions and hardware developers to help inform the development of the hardware to match the highest value applications,” he says.

In terms of SQC, a firm which aims to build the world’s first silicon quantum computer, and in which Australia’s CBA is a stakeholder of, the latter’s Rajasingham says that the goal is to have a 10-qubit fully-entangled device by the year 2022.

From CBA’s perspective, he feels it will reach “phenomenal targets” due to its involvement with SQC and also the quantum simulator that QxBranch deployed on-premise to experiment and work with.

“On a silicon architecture, we believe we can solve a huge number of problems…because silicon is a very well-understood material. There are a lot of nuances in keeping a quantum state and many of those are significantly simplified—although there’s a lot of work to be done—by using materials that you’re very familiar with and the industry knows a lot about,” he says.

He says that silicon in a quantum state allows for the longest coherence time, and a high-quality qubit gate operation time, which is the amount of time that effective operations can be conducted before the state collapses. “How long is the quantum state available for, how long is it effective for or how much processing can you do? All these characteristics are really important when you look at a quantum computer,” he says.

Working with silicon also allows for scalability, which he notes, is one of the issues quantum teams all over the world are trying to tackle.

“The quantum state is a very fragile state to maintain, so the more that you know about the material that you’re keeping the state in, the better off you are,” he adds.

In terms of coherence times, there are many ways the architecture can help prolong control of the qubits from using nanowires specifically designed to control them through the materials that are used, through to the dilution fridges to cool everything down to just about absolute zero. This is a new era of quantum computing called the noisy, intermediate-scale quantum computer (NISQ), as John Preskill, the Richard P. Feynman professor of theoretical physics at the California Institute of Technology, describes it.

In a keynote speech presented in early December 2017, he said today’s best hardware, referring to superconducting circuits or trapped ions, the probability of error per two-qubit gate is about 10^-3, and the probability of error per measurement is about 10^-2 or better for trapped ions.

“Naively, we cannot do many more than 1,000 gates—and perhaps not even that many—without being overwhelmed by the noise. Actually, that may be too naive, but anyway the noise limits the computational power of NISQ technology,” he said.

Not So Easy

Any advancement in the quantum space is sure to generate excitement about how the industry is moving closer and closer to unlocking QC’s true potential. But there are still many things to iron out and many questions that are left to be answered. The fact is that a true commercial quantum computer is still very far away.

It’s just as theoretical physicist Feynman once put it: “Nature isn’t classical, dammit, and if you want to make a simulation of Nature, you’d better make it quantum mechanical, and by golly it’s a wonderful problem because it doesn’t look so easy.” 

We’re just at the genesis of quantum computing, but after a long period of fits and starts, its evolution is now happening faster than many had expected. Now is the time for firms to start paying attention.