What does quantum computing mean for the safety of data? Consider this.

A computer with 1 teraflop of power, capable of one trillion operations per second, would need around 300,000 years to breach the widely used RSA 2048 bit encryption algorithm.

A quantum computer, however, could achieve the same feat in just 10 seconds.

On the brighter side, using quantum technology for encryption would allow you to send confidential data to intended recipients in a way that is virtually unhackable — even by another quantum computer.

The potential implications of this computing power are vast for many areas of modern society, but the impact for the banking industry, with the colossal amounts of confidential data it handles and financial analysis it produces, is profound.

That is, when the quantum revolution actually arrives.

After decades of exploration and research, quantum computers do exist, but they are few and not yet quite effective or powerful enough to replace classical computers.

Technology giants are making strides in the quantum arms race, as banks and other companies in various industries gear up now for this new computing era. And the tipping point may be upon us as Google published a paper in Nature in October showing that its Sycamore quantum processor completed a complex calculation that would have taken the world’s most powerful classical supercomputer, the Summit, 10,000 years to solve. This is referred to as “supremacy” in quantum computing, and has been a long-sought after milestone.

"It's going to be a game changer," said Tom Pickering, Scotiabank programmer, Global Wholesale & Risk Technology. "It's not quite there yet, everybody is running on simulators. As soon as one becomes real, it's going to be a must-have in financial services. Irrespective of whether it's three years, five years, 10 years. If you haven't got the skills in-house to be able to exploit it, you're going to be left behind."

A regular, or classical, computer uses a binary system, where information is stored in bits that encode either a 0 or 1, electrical currents that are on or off, according to Shohini Ghose, professor of physics and computer science at Wilfrid Laurier University.

Shohini Ghose, a professor of quantum physics and computing at Wilfrid Laurier University, says quantum encryption is furthest along in the space in terms of viability.

Quantum computing uses quantum bits, or qubits, which have a more fluid identity and can represent endless combinations of 0 and 1 all at the same time — a fundamental principle known as "superposition,” said Ghose.

"What quantum computing allows us to do is to move away from the binary and think more broadly on a spectrum," she said during a recent Scotiabank town hall. "So that you can think of quantum bits as having some probabilities of things… It's this fluid identity, that allows us actually more knobs. We can do computing in a much more efficient manner."

Another unique — and counterintuitive — property of quantum physics when applied to encryption is that in theory it could limit transmission of data to only those who have been intended to receive it.

Current encryption uses complex math problems to create a key to lock, and unlock, top-secret information so it can be transmitted securely.

Quantum encryption uses the properties of physics instead of math and the key is encoded into particles transmitted between two parties. Thanks to the fundamental quantum mechanics principle of uncertainty, the mere observation or attempt to copy or measure these particles changes them and any attempt by an outsider to do so would be detected immediately.

"They can't just secretly take all the information and go away and decrypt it ... those who are creating the key will always detect this eavesdropping, because of that disturbance,” said Ghose. “As soon as they know somebody is trying to sort of copy the data or harvest it or sneakily look at part of it, immediately, they stop the key process and restart it again.”

Quantum computing still facing hurdles to scale

While IBM, Google, the Canadian company D-Wave and others already have quantum computers they are still in their infancy as there are significant challenges to scaling the technology up for the broader market.

For example, they must be kept at extremely cold temperatures in order to operate.

"Our laptops overheat, that's a big problem,” Ghose said. “Well, (current) quantum computers overheat at a whole different level. They have to operate at temperatures that are like 200 times colder than outer space... That's the level of challenge."

Quantum encryption, however, does not require full-scale quantum computing and this technology is the furthest along in the space in terms of viability, said Ghose.

The ability to transmit data only for intended eyes is “extremely powerful,” said Daniel Moore, Scotiabank’s Chief Risk Officer.

“That's a very important use case. That can't be deployed at scale yet. But, in my opinion, its a question of when that happens and not if."

With that big opportunity comes risk as well.

“Even if quantum computers are not developed for the next decade or so, when those computers are developed, they can actually hack most of the current encryption,” said Ghose. “If the data is saved in encrypted form today, then it can still be hacked when the computers are ready in the future. That's something that we have to think about now.”

Moore says the bank is already starting to formulate its defensive strategy, looking to protect data that will still have value when quantum technology fully catches up.

“Our data has a lifetime, it decays in its value,” he said. “Five years from now, there's no point in knowing what the balance of my deposit account is. Doesn't help you very much. But knowing my date of birth, social insurance number, that probably will not change in five years’ time. That data has a persistence that may be of a significant value.”

One option is to move toward quantum-safe encryption methods.

“Those are classical methods, but they're demonstrably safe from quantum decryption,” Moore said. “We're understanding what the solutions to the problem are, and we’re starting to have that dialogue about what our defensive strategy would be. What are your crown jewel assets that you need to protect first, with the highest defences, and how would you do that?”

'A thousand-fold-plus speed increase'

Meanwhile, Scotiabank is also proactively exploring quantum computing’s potential applications to trading.

In August, Toronto-based photonic quantum computing company Xanadu announced the results of a proof-of-concept technology collaboration between Scotiabank and another Canadian Bank. The aim of the collaboration was to discover how to increase the speed and accuracy of computations in the trading products space.

Financial institutions use a gigantic amount of resources and time to price portfolios of trading products to gauge different possible market scenarios. They use a method that randomly samples probability distribution known as Monte Carlo estimation to model and price derivatives, often running in large data centres of parallel central processing units (CPUs) and graphic processing units (GPUs).

The proof-of-concept used an algorithm developed by Xanadu called quantum Monte Carlo and a software suite to simulate it on various trading products.

"It's a thousand-fold-plus speed increase on the most complicated calculations we do,” said Pickering about the project’s results. “And speed over repeated calculations gives us better accuracy.”

Artificial intelligence and deep learning approaches are almost as fast, but unlike quantum, they are available right now, he said. The Bank’s approach is to leverage these more tangible technologies now but make the switch to quantum once it goes live.

“It’s sufficiently different and sufficiently close that we have to keep a watch on it,” Pickering said.

Stella Yeung, SVP & CIO Scotiabank Global Banking & Markets, says they plan to do more proof-of-concept projects and have a research team focused on quantum to be ready for when commercial availability dawns.

Pickering and Yeung estimate this could be in three to five years.

“We don’t have a crystal ball,” said Yeung. “But as long as we keep the momentum (to keep up) … We will continue to do that in a measurable way.”