This story is part of CNET’s look at how the world will continue to evolve in 2022 and beyond.
After years of development, quantum computers reached a level of sophistication in 2021 that encouraged commercial customers to start dabbling in the radical new machines. Next year, the business world may be ready to welcome them more enthusiastically.
BMW is among the manufacturing giants that see the promise of machines, which harness the physics of ultra-smallness to fly past some limits of conventional computers. Earlier this month, the German auto giant picked four winners in a the contest stayed with Amazon to highlight ways new technology could help the automaker.
The automaker found Quantum computers have potential to optimize sensor placement in automobiles, predict metal deformation patterns, and employ AI in quality control.
“We at the BMW Group are convinced that future technologies such as quantum computing have the potential to make our products more desirable and more sustainable,” Peter Lehnert, who heads BMW’s research group, said in a statement.
BMW is not alone in its determination to evaluate the practical application of quantum computers. Aerospace giant Airbus, financial services firm PayPal and consumer electronics maker LG Electronics are among the commercial companies looking to use the machines to refine materials science, streamline logistics and monitor payments.
For years, researchers have worked on quantum computers as more or less conceptual projects that take advantage of qubits, data-processing elements that can contain more than the two states that drive transistors found in conventional computers. Although they got better, quantum computers were better suited to research projects, some as basic as figuring out how to program exotic machines. But at the current rate of progress, they will soon be powerful enough to tackle computing jobs beyond the reach of conventional computers.
Like cloud computing before it, quantum computing will be a service that most corporations rent out to other companies. Teams require constant attention and are notoriously complicated. Although more work is required to realize their full potential, quantum computers are becoming increasingly stable, a development that is helping corporations overcome initial hesitation.
Georges-Olivier Reymond, CEO of the startup Paschal, says progress is changing skeptics who previously viewed quantum computing as fantasy. A few years ago, employees at large corporations would roll their eyes when he brought it up, but that has changed, says Reymond.
“Now every time I talk to them I get a positive response,” Reymond said. “They are ready to participate.”
A new client is the European defense contractor Thales, which is interested in quantum computing applications in sensors and communications. “Pasqal’s quantum processors can efficiently tackle large problems that are completely beyond the reach of classical computing systems.” Thales Chief Technology Officer Bernhard Quendt he said in a statement.
Of course, quantum computing is still a small fraction of the traditional computing market, but it is growing rapidly. About $490 million was spent on quantum computers, software and services in 2021, Bob Sorensen, analyst at Hyperion Research said in the Q2B conference held by quantum computing software company QC Items in December. He expects spending to grow 22% to $597 million in 2022 and an average of 26% annually through 2024. By comparison, spending on conventional computing is expected to rise. increase 4% in 2021 to $3.8 trillionGartner analysts predict.
The growing business activity is notable given that using a quantum computer costs between $3,000 and $5,000 an hour, according to Jean-Francois Bobier, an analyst at the Boston Consulting Group. A conventional high-performance computer hosted on a cloud service costs half a penny for the same amount of time.
Analysts say real spending on quantum computing will start as the industry tackles error correction, a solution to the vexing problem of easily perturbed qubits derailing calculations. The fidelity of a single computation step on the most advanced machines is around 99.9%, which leaves a degree of flaking that makes a raw quantum computation calculation unreliable. As a result, quantum computers have to run the same calculation many times to ensure that the answer is correct.
Once error correction is mature, the revenue generated through quantum computing will explode, according to the Boston Consulting Group. With current machines, that value will likely total between $5 billion and $10 billion by 2025, according to the consultancy’s estimates. Once the bug-fixed machines arrive, the total could go as high as $450 billion to $850 billion by 2040.
Software and services that hide the complexity of quantum computers will also drive usage. IonQ CEO Peter Chapman predicts that by 2022, developers will be able to easily train their AI models with quantum computers. “You don’t need to know anything about quantum,” Chapman said. “You just give it the dataset and it returns you a model.”
Among the signs of commercial interest:
- Automaker Daimler and oil giant ExxonMobil have hired IBM, an early proponent of quantum computing, to use some of the tech company’s 22 machines. (Big Blue will soon add Eagle, a new, more powerful device that uses a 127-qubit processor.) The companies are prime candidates for the advances in materials science that quantum computers could one day deliver. They could also use the machines to help with logistical challenges, like getting the right equipment in the field when it’s needed or the right components to the factory in time for manufacturing.
- Airbus, LG Electronics, healthcare company Johnson and Johnsonenergy company FED and the chemical giant BASF are among the industrial companies vying for time on Pasqal’s unique quantum computer. The startup is working on a relatively new “neutral atom” design that uses lasers to carefully arrange rubidium atoms and then use them as qubits to process data. Many quantum computing startups are in the US, but France-based Pasqal benefits from government programs and European customers.
- PayPal is testing a quantum anneal, a type of quantum computer suitable for solving optimization problems and made by D wave systems, to guide its fraud detection work, according to Vidyut Naware, director of AI research at the company. He also told the Q2B crowd that PayPal is running the test to get a better idea of how quantum computing can be implemented more effectively.
- Amazon Web Services offers customers access to machines from IonQ, Rigetti Computing and D-Wave Systems. It will soon add QuEra machines and Oxford Quantum Circuitsand is building his own quantum machine. Among those using its services is JPMorgan Chase, which is testing it for optimization work, such as managing its clients’ investment portfolios.
- Microsoft is working on its own approach to quantum computing, called topological qubit design. The work is not yet commercially available, but Microsoft’s Azure cloud service offers access to quantum computers and quantum computer simulators that are an important stepping stone. The company has also struck a deal with KPMG to help customers test the quantum computing water using Azure.
- Quantinuum, the company formed when the hardware manufacturer service in december to generate random numbers. That may sound obscure, but those numbers are the foundation for modern encryption. Among those testing the service are Fujitsu and axiom space. launched a
- Commissioning of quantum chemistry software Qunasys offers software to clients such as Mitsubishi Chemical CorporationCEO Tennin Yan said. Quantum chemistry and materials science could help engineers find better materials for solar panels, electric vehicle batteries, and plastics.
Quantum computers today are more of a luxury than a necessity. But with its potential to transform materials science, shipping, financial services and product design, it’s no surprise that companies like BMW are investing. The automaker will benefit from knowing better how materials will deform in a crash or training the vision AI of its vehicles faster. Although quantum computers may not produce a bounty this year or next, losing the technology once it matures comes at a cost.