Quantum scientist and author Chris Ferrie poses a seemingly simple question: how will we know when quantum computing has truly arrived?
The relentless march of digital technology has seen the number of working components in a computer double approximately every two years since the 1970s, an observation known as Moore's Law.
Transistors are tiny switches that can turn on and off to represent bits of information, ones and zeroes, the building blocks of digital data. You have billions of transistors in your smartphone. To fit more of them on a computer chip requires them to be smaller — but this trend of miniaturization cannot continue forever.
Eventually, transistors will be the size of indivisible atoms.
The quantum limit
Hypothetically, if transistors were the atomic in size, they would need to be described and controlled using quantum mechanics. This would necessitate the processing of information not encoded as ones and zeroes, but as the much more mathematically rich set of complex numbers that characterize quantum systems — counted in fundamental units called quantum bits, or qubits.
In the 1980s, it was suggested that a computing device operating according to the rules of quantum mechanics could perform certain tasks faster or in fewer steps than digital computers. In 1994, Peter Shor, then a staff scientist at Bell Labs, debuted his namesake quantum algorithm, a sequence of steps performed on qubits of data that could factor numbers with an efficiency far surpassing any known digital algorithm.
Until then, factoring large numbers was thought to be a difficult problem, a fact that lies at the core of secure public communication, such as private internet transactions. The potential to “break the internet” spurred a global race to develop real quantum computers, and cryptography unbreakable by quantum algorithms.
After several decades of progress, we are nearing a transition point where quantum computers become commercially viable.
A “quantum” Moore's Law?
As we stand on the brink of this second quantum revolution, one can't help but wonder if there could be a "quantum" equivalent to Moore's Law.
In 2019, Hartmut Neven, director of the Quantum Artificial Intelligence Lab at Google, proposed “Neven's Law,” suggesting that the number of qubits in quantum computers appeared to be doubling at a rate similar to Moore's Law.
However, the challenge is that physical systems used to encode qubits are inherently prone to errors; qubits counts alone, therefore, do not correlate with computational power unless errors can be kept in check.
The holy grail of quantum computing is the development of a device that can carry out instructions to any desired precision with little overhead — a fault-tolerant quantum computer. In theory, this requires extremely low error rates and active error correction.
While we are not there yet (as of mid-2024), demonstrating the power of quantum computing may still be possible with error-prone devices.
The big question everyone wants to know the answer to is this: how will we know quantum computing has “arrived?”
Quantum supremacy
In 2012, John Preskill, a theoretical physicist at the California Institute of Technology, introduced the term “quantum supremacy” to describe the point at which a quantum computer can solve a problem that is essentially infeasible for classical computers.
Preskill's vision was ambitious. “In a race, a horse has an advantage if it wins by a nose. In contrast, the speed of a quantum computer vastly exceeds that of classical computers, for certain tasks,” he explained.
Quantum supremacy, in this analogy, is akin to a sports car effortlessly outpacing horses in a race.
While factoring an extremely large number would be a clear demonstration of quantum supremacy, Preskill suggested that sampling problems — producing random numbers with a particular bias — might be an easier target.
The race went into overdrive.
The Google announcement
In 2019, Google's researchers announced that they had achieved quantum supremacy with their Sycamore processor, a 53-qubit quantum device.
Sycamore completed a specific task in 200 seconds that (they claimed) would have taken the most advanced classical supercomputers thousands of years to finish.
Critics pointed out that the task was artificially constructed and had no practical applications, thus questioning the real-world relevance of this achievement. Other counter-claims directly questioned the comparison, suggesting the feat could be achieved in a few days on a supercomputer.
From supremacy to advantage
Google’s result demonstrated that a quantum computer could indeed outperform classical computers in specific tasks, even if those tasks were not practically useful.
This realization shifted the focus from the concept of quantum supremacy to the more pragmatic goal of quantum advantage — achieving a tangible, practical benefit from quantum computing in solving real-world problems.
In Preskill’s analogy, we would be very happy to have bet on the horse that “merely” wins the race rather than demanding that having dominated it.
Quantum technology has no standardized definitions, so the term “quantum advantage” can have several meanings, including being synonymous with “quantum supremacy” or referring to “supremacy but for a practical problem.”
However, it’s more enlightening to use the term to mean something more general: Quantum advantage is achieved when a quantum computer is preferred in performing a task of practical relevance.
The definition does not require the “domination” demanded by supremacy and also allows the advantage to be more than time-saving — a quantum computer could be preferred because it consumes less energy or is more accurate, even if it does not provide a speed-up.
On the other hand, the requirement of “practical relevance” is stronger than one might expect. As far as we currently understand, quantum algorithms to solve real-world problems require fault-tolerant operation.
A quantum future
The ultimate goal is to reach quantum advantage, where quantum computers can solve meaningful problems more efficiently or effectively than classical computers.
The practical realization of quantum advantage necessitates collaboration across disciplines and substantial investment in research and development. Governments, academia, and industry must collaborate to overcome these challenges and unlock quantum computing's full potential.
As we navigate this uncharted territory, the pursuit of fault-tolerant quantum computers remains a driving force, propelling us toward a future where quantum technology transforms industries and society as profoundly as classical computing has over the past half-century.
Chris Ferrie is an associate professor at the University of Technology Sydney and Centre for Quantum Software and Information. He is the author of the successful Baby University series, including the breakout success Quantum Physics for Babies.
Comentarios