Quantum computing is believed to harbor the potential to absolutely revolutionize, or disrupt, a variety of industries and sectors. Large, fault-tolerant quantum computers may be able to perform complex calculations that are thoroughly impractical, if not outright impossible, for the world’s most powerful supercomputers now and into the future. Although the technology is still in its early stages of development, enterprises across the spectrum are exploring its potential applications, not to mention its potential implications.

What is the Timeline for Quantum Applications?

The development of quantum computing applications is heavily dependent upon the state of current hardware. Theoretical research continues, but experimentation and verification require much larger devices with much lower error rates than is currently available. Nonetheless, current technology and classical simulation allow the development of proofs-of-concept, benchmarking versus classical alternatives, and feasibility determinations. Enterprise R&D, therefore, has not one, but three distinct timelines to consider.

In the near-term, which is not strictly defined, but zero-to-five years is a common range, enterprises are looking to find innovative ways to harness the technology that is available, despite the previously-mentioned shortcomings. Modern-day quantum computers are small and error-prone, but if there is a computational advantage to be found, there is an ongoing race to find it. Such an advantage has not yet been found, but the quest for it has led to hybrid classical-quantum approaches that leverage high-performance computing (HPC) to maximize performance, as measured by such metrics as runtime and accuracy.

In the intermediate-term, generally considered within the next 5-10 years, enterprises must consider that the technology will greatly improve and error rates will significantly decrease. However, problem sizes will remain relatively small. In such a scenario, computational advantages might already be realizable over classical systems, but the largest problems might still be out of reach.

Finally, in the long-term, which will likely be at least a decade from now, enterprises can imagine the existence of large, fault-tolerant quantum computers that are able to perform calculations that are classically intractable. Problems that cannot be solved today may routinely be solved. Problems for which solutions can only be approximated today may be solvable with greater precision. These are the ultimate goals of today’s research programs.

Top Applications of Quantum Computing

Quantum computer applications fall into a small number of classifications. The counts and names of these classes vary from source to source, but not significantly. The Quantum Insider, for example, identifies five of them, specifically:

  • Finance
  • Machine Learning (ML)
  • Material Science
  • Natural Language Processing (NLP)
  • Optimization

 The number of quantum computer use cases is noticeably larger, but these potential uses all fit into one of the aforementioned classifications. These lists have more variation, in part due to the challenge in following all of the research transpiring around the world. Consolidating lists from Analytics India Magazine and Built In, for example, non-exhaustively identifies a dozen common inclusions:

  • Artificial Intelligence and Machine Learning
  • Better batteries
  • Cleaner fertilization
  • Computational chemistry
  • Cybersecurity and cryptography
  • Drug design and development
  • Electronic materials discovery
  • Financial modeling
  • Logistics optimization
  • Solar capture
  • Traffic optimization
  • Weather forecasting and climate change

Quantum Computing Optimization Use Cases

As previously noted, optimization problems constitute one of the major classifications of potential quantum computing solutions. Research is progressing across multiple modalities, such as superconducting devices and quantum annealers, but neutral atoms have been found to be a natural fit for this class of problems.

 To superficially understand why, we can look at any visualization of a graph problem, and then we can look at an array of neutral atoms. The nodes, or vertices, of the graph are represented by the individual atoms, held in place, micrometers apart, by optical tweezers. The edges, or lines, of the graph are not visible, but can be imagined as the interactions between neighboring atoms. 

 Neutral atom quantum computer use cases are not limited to the following, however these are some of the most commonly-identified probable applications for this specific modality:

  • Antenna placement (maximum independent set)
  • Minimum vertex cover
  • 5G networks (Minimum dominating set)
  • Graph coloring
  • Weighted maximum independent set
  • Quadratic unconstrained binary optimization (QUBO)
  • Integer factorization
  • Portfolio optimization (maximum clique)‍
  • Optimizing store locations (maximum independent set)

 Quantum Computing Research Use Cases

Because of the way this specific modality works, potential neutral atom quantum computing solutions always mention maximum independent set (MIS) problems. Optimization problems are mentioned, in general, but always MIS problems specifically. For some of the latest research, be sure to check out Quantum Optimization for Maximum Independent Set Using Rydberg Atom Arrays by researchers from the Harvard-Smithsonian Center for Astrophysics, Harvard University, and the University of California Berkeley; Quantum Optimization with Arbitrary Connectivity Using Rydberg Atom Arrays by researchers from QuEra Computing, Harvard University, the University of Innsbruck, and the Institute for Quantum Optics and Quantum Information; and Industry applications of neutral-atom quantum computing solving independent set problems by a team of researchers from QuEra Computing.

Neutral atom quantum computing applications are not limited to optimization problems, however. To learn about potential Machine Learning applications, download a copy of Quantum Reservoir Computing Using Arrays of Rydberg Atoms by researchers from Harvard University and IBM Quantum. This particular paper proposes a quantum implementations of a recurrent neural network (RNN). Also, to learn about ongoing research into physics problems, have a look at Probing topological spin liquids on a programmable quantum simulator by researchers from Harvard University, QuEra Computing, the University of Innsbruck,  the Austrian Academy of Sciences, the Institute for Advanced Study, and Massachusetts Institute of Technology.

These researchers, as well as other researchers from around the world, are able to experimentally test their theories using QuEra’s quantum cloud computing system. Affectionately known as “Aquila,” this 256-atom quantum computer is accessible through AWS Braket, qBraid, and a premium service mode. Neutral atom quantum computers may also be leased for on-premise use.

‍