While 'quantum advantage' is still a few years away, the technology has clear benefits for certain classes of computation problems. Here are four important types named by experts.
Long considered the stuff of science fiction, quantum computing is on a steady course to deliver tangible benefits for select industries and distinct applications, thanks to rapid technology advances and more available and affordable means of access.
Unlike classical computing, which processes binary data (0s and 1s) in a sequential fashion, quantum computing leans into physics principles to allow for potentially faster and more efficient processing of complex problems, many of which are not solvable with the traditional approach. At the core of quantum computing is the qubit, which can represent 0s, 1s or both. Qubits can exist in multiple states simultaneously -- a physics principle known as superposition. Qubits can also be entangled, meaning they become intrinsically linked in a quantum state.
Both concepts are central to quantum computing’s primary advantage: the ability to perform probabilistic calculations in ways that are fundamentally different -- and potentially faster, cheaper, and more accurate -- than any known classical methods. Quantum's promise becomes increasingly important for certain kinds of optimization and factoring calculations as achieving sufficient performance gains on classical computers gets more complicated and, thus, more expensive and harder to realize, according to Bob Sorensen, chief analyst for quantum computing at Hyperion Research.
"Quantum offers accelerated capabilities for some very important, narrow classes of applications," he said. "You can explore a richer range of fidelity and examine a larger set of variables through superposition and entanglement."
The different ways quantum computers can handle calculations are key to many of their advantages over traditional computers.
Quantum utility, not yet quantum advantage
The category remains relatively immature, despite significant technological advancements, including in areas such as fault tolerance and error correction, which are intended to drive accuracy and reliability. There has also been a wave of startups and big-name vendors introducing cloud-based quantum computing capabilities, along with software, toolkits and platforms for developing, simulating and running quantum algorithms.
This steady progress is moving the needle a bit on the viability of quantum computing for applications beyond purely experimental ones or those involving a narrow slice of academic use cases. For now, experts are projecting a 2030 timeframe for the highly anticipated quantum advantage, which is the ability of quantum computers to solve problems that are beyond the scope of classical computers -- and in a way that is cheaper, faster or more accurate.
In the interim, however, major players like IBM contend we have reached the era of "quantum utility," when a quantum computer and quantum algorithms can reliably and accurately tackle particular problems more efficiently and effectively than is possible with traditional, brute-force classical computing. A group of MIT researchers, in partnership with Accenture, is also taking a more progressive view of when quantum computing can deliver benefits. As part of the framework they use to evaluate quantum computing's potential, the partners have established the quantum economic advantage benchmark for determining when a particular problem can be solved more quickly with a quantum computer compared with a similarly priced classical platform.
"There's a crossover for when quantum or classical computing is faster, and generally the problem needs to be sufficiently big with exponential algorithmic gains and large data sets," said Jayson Lynch, a research scientist in the FutureTech Lab at MIT. "Organizations have to understand not just their use case, but what computational problem they are trying to solve."
Quantum computing's key benefits
As the category evolves, quantum computing is not expected to serve as a direct replacement for classical computing, but rather as a complementary option for delivering substantial but highly specific advantages for certain types of problems. Here are four of the most important ones.
Complex systems simulation
Quantum mechanics is well suited for modeling processes and systems in nature that can't be accurately or adequately handled within the limitations of classical computing models. Traditionally, scientists have had to rely on physical experimentation, lab testing and limited simulation to explore how such systems behave. Quantum computing's propensity for exponential modeling and calculation gives it a leg up for discovering and simulating chemistry, material properties and high-energy molecular behavior.
"A lot of natural systems and incredibly dynamic systems are really difficult if not impossible to do with a classical approach," said Jeannette Garcia, senior research manager for quantum applications and software at IBM. "With quantum, there are a lot of knobs you can turn from a computational standpoint."
Search and optimization
Identifying optimal material properties or gaming a large-scale supply chain scenario involves searching for the best solution among endless permutations and possible combinations. Classical computing takes a linear approach, executing one complex calculation after another, which is limiting, given the range of potential variables. In comparison, quantum computing can explore many variables simultaneously, which can quickly narrow the range of possible answers.
"Every business has an optimization opportunity, whether that's the layout of a factory to move materials more efficiently or in the financial services sector for investing client money," said Hyperion's Sorenson. "You don't get the perfect answer out of quantum today, but you may get a better solution than out of a classical counterpart, and that's attractive."
Classification and anomaly detection
Quantum computing's ability to perform complex, multivariable calculations simultaneously gives it certain speed and scale advantages for training machine learning (ML) and AI algorithms. As a result, quantum computing excels at anomaly detection, which is crucial for efficient AI and ML processing. "There are certain problems in optimization and AI/ML, where classical computing algorithms look at data and see randomness while quantum algorithms can find patterns in what looks like random noise," said Scott Crowder, vice president of quantum adoption at IBM.
Quantum computing opens up a chance to look at all the problems we can't solve with classical computing and think about them in a different way.
Scott CrowderVP of quantum adoption, IBM
Factoring and cryptography
The most well-known quantum algorithm was developed in the 1990s by MIT mathematician Peter Shor to factor large composite numbers more efficiently. While no existing quantum system can run it at scale, the algorithm is theoretically capable of breaking widely used encryption -- and remains one of the best examples of potential quantum advantage. Some businesses and organizations already use so-called post-quantum cryptography methods (several of which NIST recently standardized) to protect sensitive data from potential decryption attacks by future quantum computers.
Eyeing new horizons
While quantum computing won't map to every problem, its robust benefits should be a wake-up call for IT leaders to start thinking about the technology and keep an open mind. "Quantum computing opens up a chance to look at all the problems we can't solve with classical computing and think about them in a different way," Crowder said.
Beth Stackpole is a veteran business and technology reporter who has spent 30-plus years writing for a variety of publications and websites, including Informa TechTarget, Computerworld, CIO, eWeek and Automation World.
