Framework from MIT can help businesses anticipate when Quantum Computing might be useful.
Once-obscure technologies can emerge in an instant, causing companies to scramble to figure out how to effectively leverage them for competitive advantage.
Table of Contents
Quantum computing, which most people can’t define and still don’t fully comprehend, could be the next obscure technology to have a seismic impact on business. Quantum computing uses quantum mechanics laws to simulate and solve complex problems that are too difficult for today’s classical computers.
In some cases, quantum computers can solve large-scale problems much faster than their classical counterparts. Simulating matter’s behavior, analyzing compounds to create new drugs, optimizing factory floors or global supply chains, and detecting fraud and risk patterns in financial transactions are some examples.
McKinsey estimates that 5,000 quantum computers will be operational by 2030, but that the hardware and software required to handle the most complex problems will not be available until 2035 or later. However, organizations must begin planning now for how they will use technology to solve real-world business problems. According to a November 2022 report, some companies expect to invest more than $15 million per year in quantum computing.
A group of MIT researchers, in partnership with Accenture, has developed a framework to help tech-savvy executives start to evaluate the potential of quantum computing for problem-solving at their companies.
“Implicitly, there’s a race going on between the classical computer and quantum computer. For each type of question you want to solve, you want to know which type of computer will win so you can take the best advantage of it,”
Neil Thompson, a research scientist at MIT Sloan and the MIT Computer Science and Artificial Intelligence Laboratory, said at the 2023 MIT Initiative on the Digital Economy annual conference.
Thompson is a co-author of “The Quantum Tortoise and the Classical Hare: A Simple Framework for Understanding Which Problems Quantum Computing Will Accelerate (and Which It Won’t),” along with Sukwoong Choi, an assistant professor at the University of Albany and a digital fellow at the MIT Initiative on the Digital Economy, and William S. Moses, an assistant professor at University of Illinois Urbana-Champaign.
“This framework provides a way to analyze the potential impact of switching to quantum computing before making the investment,”
The researchers’ takeaway is that small to moderate-sized problems, the most common types for typical businesses, will not benefit from quantum computing. Those trying to solve large problems with exponential algorithmic gains and those that need to process very large datasets, however, will derive advantages. “Quantum computing is not going to be better for everything, just for some things,” Thompson said.
An overview of quantum computing
The idea for building a system that leverages physics principles to simulate problems too difficult to model with traditional digital systems was first proposed in the 1980s. The concept was buttressed by MIT mathematician Peter Shor, who developed the first well-known quantum algorithm for breaking encryption in the 1990s.
Unlike today’s computers and supercomputers that use binary electrical signals to represent ones or zeros, quantum computers employ quantum bits (qubits), which are subatomic particles. When managed properly, qubits can represent combinations of both ones and zeros simultaneously. The more qubits, the greater potential for large-scale compute power for problem-solving.
When quantum computing will be useful
The crux of the framework in “The Quantum Tortoise and the Classical Hare” is that classical computers (the hare) are generally faster than quantum computers (the tortoise), but require more steps to complete a task — which the researchers liken to taking an inefficient path from point A to point B.
Quantum computers, with their ability to run more efficient algorithms, have the potential to take a more direct path — but because they have slower processing speeds, it may take longer to solve the problem. The framework developed by the researchers aims to assist businesses in determining whether the shorter route or the faster computer is more valuable, depending on the problem at hand.
Scientists are working to achieve quantum advantage, or the ability to use quantum computers to solve problems that classical computers cannot. (It is estimated that some companies will achieve the quantum advantage by 2030.)
However, the researchers believe that the emphasis on quantum advantage overshadows the utility of quantum computers as they become cost-competitive with traditional computers. To make up for this, their framework establishes the benchmark of quantum economic advantage, which occurs when a specific problem can be solved faster with a quantum computer than with a comparably priced classical computer.
To determine the quantum economic advantage, business and technology leaders will have to consider two conditions:
- Feasibility, meaning whether a quantum computer exists that is sufficiently powerful to solve a particular problem.
- Algorithmic advantage, meaning that a quantum computer would be faster at completing a particular task compared with a comparably priced classical computer.
The overlap between the two is the quantum economic advantage. Thompson advised businesses to consider the speed of the computer versus the route. “Think of it like a race in getting from point A to point B, and the algorithm is the route,” Thompson said. “If the race is short, it might not be worth investing in better route planning. For it to be worth it, it has to be a longer race.”
Other things to consider about the current state of quantum computing
Although it is still early days, quantum is heating up. Although quantum computing is still in its early stages of development, the landscape is heating up. IBM released Osprey, a 433-qubit machine, last year and plans to build a 100,000-qubit machine within the next ten years. Google’s goal is to have a million qubits by the end of the decade. D-Wave Systems, IonQ, Rigetti Computing, Honeywell, Microsoft, Intel, and PsiQuantum are among the other companies offering quantum computing services in the cloud. According to Fortune Business Insights, the quantum computing market will grow from $928.8 million this year to $6.5 billion by 2030, representing a compound annual growth rate of 32.1%.
The challenges of development, cost, and talent remain. Companies are still figuring out how to scale the number of physical qubits that can be built into quantum computing systems, as well as how to optimize how the various qubits interact with one another as horsepower scales. A lot of research is currently being done to reduce the error rates, or noise, in quantum computing. The technology is also expensive because the systems require complex cooling technologies to protect the qubits.
The skills gap is another problem: Subject-matter experts are hard to come by outside of research and academic circles. McKinsey predicts that by 2025, fewer than half of quantum jobs will be filled, which is a major barrier to adoption.
Benefits will come on a continuum. Quantum computing becomes more attractive when the quantum algorithm is exponentially faster or significantly better than the classical computing option, or if the problem size being tackled is larger than the speed differential between the two. Given the current state of quantum technology, it will be useful sooner for small-scale problems whose solutions offer smaller benefits and will only later be viable for solving more complex problems that promise larger benefits.
Source: MIT