Technologies and trends
shaping the future of commerce
USE ARROW KEYS OR SWIPE
Q4 2019 Edition - Article 01
Imagine it was possible to solve complex mathematical problems exponentially faster than today’s most advanced supercomputers. What kinds of problems would we tackle? How might this expansive computing power impact our business and our industry? The emerging field of Quantum Computing may enable this very reality far sooner than anyone expects.
In a classical computer, microprocessors with integrated circuits containing billions of tiny transistors serve as the basis for computation. Each of these billions of transistors can be set to one of two states, ON (1) or OFF (0), creating binary digits or bits. These bits then enable the calculations and processing required to bring everything from pocket calculators to virtual reality headsets to life.
In quantum computing, bits are replaced with “quantum bits” or simply qubits. Qubits are created by isolating subatomic particles—a feat in and of itself. Once these qubits are isolated, researchers can take advantage of quantum mechanical properties to unlock exponentially more computational power than can be delivered by a classical bit.
In this position paper, we will discuss the
key characteristics or mechanics of quantum computing that enable this computational power, the use cases quantum computers could address, and the timelines against which success is measured.
vs quantum computers
Principles of Quantum Mechanics
THREE PRINCIPLES OF QUANTUM MECHANICS
A detailed overview of quantum mechanics is beyond the scope ofthis paper; however, it will be helpful to touch on three underlying principles of quantum computing:
Unlike bits that may be placed into either one of two states (1 or 0), qubits can be placed in an infinite number of states including 1, 0, and all of the possible intervals between—at the same time. This is known as Superposition. As an analog, imagine a bit is represented by a coin with a clear head and tail. A qubit then is represented by the sphere that emerges when the coin is spun on its axis. With some probability, the spinning coin is a mix of both heads and tails. Only by interfering with/observing the spinning coin do we reveal its actual state.
1 2 3
When pairs or groups of quantum particles are entangled, they behave in ways that are perfectly correlated with one another—even when separated at great distance. Entanglement allows qubits to communicate with each other, enabling the effects of superposition and resulting in the exponential scaling of computing power.
In classical computing, each bit allows the computer to be placed in a total of n² states (where n is the number of bits). Thus, a 64-bit computer can be placed in a total of 4,096 states. Conversely, each added qubit in a quantum computing system allows for the system to be placed in 2ⁿ states (where n is the numberof qubits). As a result, a 64-qubit computer can be placed in 18,446,744,073,709,551,616 (more than eighteen quintillion) possible states.
POSSIBLE USE CASES
Quantum computers are not a replacement for classical computers. They may, however, address a very specific set of questions that classical computers are ill-equipped to tackle:
Not surprisingly, quantum computers may make it possible to model the behavior of quantum systems, i.e. systems in nature that have quantum properties. Simulating molecules at the atomic level, for example, is largely beyond the capabilities of a classical super-computer. This kind of modeling could have profound implications for the testing and development of new medicines, materials, and chemicals.
In financial services, at least one investment bank has explored the idea of modeling investment portfolios based on naturally occurring phenomena with quantum properties. This work is of course highly theoretical—for example, using models of the sun’s surface to develop hypotheses about the volatility of investment risk—and potentially decades away from mainstream application.
MODELING QUANTUM SYSTEMS
LARGE OPTIMIZATION PROBLEMS
Optimization problems are another often identified use case for quantum computing. Optimization problems attempt to improve the overall performance of a system across a number of variables. For example, quantum computing could maximize the number of convenient airline routes while minimizing flight time, fuel costs, personnel, and carbon footprint.
In digital commerce, retailers could realize meaningful improvements to the bottom line from better optimizing shipping and distribution operations. Consider, for example, all of the factors impacting the efficiency of Amazon’s fulfillment network—location, purchase volume, parcel dimensions, fuel costs, vehicle speed, to name a few—and it’s easy to see where classical computing models reach their limit.
A quantum computer, developed by IBM, is reminiscent of the early mainframe computers used in the middle of the 20th century. Most of the structure shown here is required to isolate the qubits from interference and maintain a temperature just a few degrees north of absolute zero (0˚ Kelvin).
Optimization could also be applied to transaction and settlement behavior. For example, at least one financial institution is exploring the use of quantum computing to optimize securities transaction settlements. This could be especially useful in environments with varying credit, collateral and liquidity constraints.
A similar idea would be to use quantum computing to identify the optimal source of funds for any given transaction. Using data about where the consumer is, what they are purchasing, the balances of the consumer’s accounts, the costs and benefits associated with a given purchase instrument (rates, penalties, impact on credit score, rewards, etc.), it would be theoretically possible to build an optimization program that chooses the best possible financial instrument for a given purchase. With the power of 5G, this kind of analysis and information sharing could theoretically happen in real time at the point of sale.
Lastly, we note that quantum computing may be used to speed up the training of Artificial Intelligence (AI). Both Machine Learning and Deep Learning use complex optimization functions to evaluate probabilities and determine an output. This computationally intense process, it is believed, can be accelerated and improved with quantum computing.
PRIME NUMBER FACTORING
One use case particularly well-suited
for quantum computing is factoring large numbers. Factoring is a mathematical tool fundamental to cryptography. For small numbers, the process is relatively straightforward and can be done with pen and paper, but things get quite a bit more complicated as the numbers get larger.
Using a quantum computing algorithm (Shor’s Algorithm), it has been demonstrated that a quantum computer of sufficient power could factor very large numbers in relatively short periods of time. This has raised some alarms across the cryptography community, as, in theory, quantum computing could be used to undermine at least some of the more common cryptographic algorithms used to secure online communications.
As a result, efforts have been underway for several years to develop post-quantum or quantum-proof cryptography. These new cryptographic functions are powerful enough to withstand or at least deter malicious intent that may arrive from quantum computing in the near term.
While quantum computing could introduce challenges for encryption and cryptography, “quantum key distribution” or “quantum cryptography” could theoretically enable even greater encryption technologies.
One scenario for quantum cryptography takes advantage of the fact that quantum particles exist in multiple states at the same time (superposition) and do not reveal their actual state until they are observed. In addition, the conditions under which the observation is being made determine the ultimate state of the particle.
The implication of these quantum mechanical properties is that even if someone was able to observe an encrypted message, they would need to do so under very specific conditions. As these conditions could be changed for every message, the costs associated with breaking a message’s encryption would far outweigh the value of the data in the message itself.
The second scenario considers a mechanism to protect against unwanted attempts to eavesdrop upon encrypted messages. The underlying idea is to use the property of entanglement to immediately recognize any attempts to “observe” the message—a quantum tripwire, so to speak.
THE END OF CLASSICAL COMPUTING?
It should be emphasized that reaching quantum supremacy will not make classical computers obsolete. For one, classical computers are acting as both input and output mechanisms. Additionally, as previously discussed, quantum computers are theorized to be useful for only a limited (albeit interesting) set of use cases.
Despite recent advances, quantum computing as a commercial application appears to be in a still very nascent state. Google, IBM, Intel, and others continue to announce higher and higher qubit quantum computers and processors, and we expect to see accelerating progress in the field. Still, most experts believe we are more than a few years, if not decades, away from producing quantum computers with any material commercial utility.
“A classical computation is like a solo voice—one line of pure tones succeeding each other. A quantum computation is like a symphony—many lines of tones interfering with one another.”
— Seth Lloyd
Our product and innovation teams are continuously exploring how emerging technologies and computational paradigms may support efforts to deliver safe, simple, and smart solutions. In the specific case of quantum computing, we have taken and continue to take steps to future-proof our network security and evaluate new analytic tools that might further inform our understanding of commerce and payment trends.
THE QUEST FOR QUANTUM SUPREMACY
At some point, researchers will aim to produce quantum computers with certain technical characteristics (e.g., sufficient number of qubits, low error rates, and shallow computational
depth) to achieve what is known as Quantum Supremacy. Quantum supremacy is the point at which a quantum computer will outperform a classical computer for certain, well-defined mathematical tasks.
In September 2019, Google published a draft paper to NASA’s website claiming to have achieved quantum supremacy. Using a quantum computer—specifically developed for the application in question—Google’s researchers were able to perform a mathematical task predicted to take almost 10,000 years to complete with a classical computer in just over
3 minutes. While some experts have discounted the magnitude of the feat due to the contrived nature of the problem, the results certainly encourage continued exploration around the potential of quantum computing.
A mathematical task predicted to take almost 10,000 years to complete with a classical computer was performed in just over 3 minutes.
explores the technology and trends shaping the evolution of commerce in today’s digital-first world.
For more information
on this topic or other digital topics of interest, please contact the Digital Future team at: email@example.com or your Mastercard account representative.