Quantum computing offers the promise of being the next big technological breakthrough. However, the many companies attempting to develop commercially viable quantum systems are pursuing a host of different technological approaches. While some of these may pan out, others may not. And while some could reach a useful level of accuracy soon, others may take considerably longer.
Let's look at four of the main quantum computing technologies and the companies pursuing them.
1. The superconducting qubit approach
All quantum computers are built around qubits -- their fundamental units for holding and processing information. But just how one goes about creating and manipulating those qubits is where the variation gets wide.
One of the most common approaches being pursued uses superconducting qubits: electronic circuits made of superconducting materials that must be brought to temperatures near absolute zero, and that are controlled using microwave pulses. These circuits can be manufactured in traditional semiconductor fabs, and the qubits don't require complex laser setups to hold them in place.
The biggest strength of the technology is its gate speed, which is much faster than some of the other techniques being pursued. However, its fabricated qubits are more susceptible to being influenced by external forces, which leads to more errors in their calculations. They also require expensive dilution refrigerators to keep the entire system cooled to the temperatures where the qubits' behavior is governed by quantum mechanical principles.
Among the companies pursuing this technology are IBM (IBM 3.25%) and Rigetti Computing (RGTI 2.06%). Rigetti has had documented issues with delays and accuracy, and seems further behind many of its peers in the race to deliver a useful system. While IBM's systems also lag in fidelity measurements, it is taking some novel approaches to solving the error-correction problem, such as c-couplers and real-time error-correction decoders, that look promising.
Image source: Getty Images.
2. The trapped-ion quibit approach
The trapped-ion approach uses individual charged atoms (ions) to make each qubit, holding them in electromagnetic fields and manipulating them with lasers. These qubits can hold their quantum states (coherence) for much longer than superconducting qubits. The big advantage of technology is that it has been developed to the point where it offers the most accuracy so far.
The biggest downside to the technology is that it is exponentially slower than the superconducting qubit approach. It also generally requires complex setups of lasers and mirrors to keep the ions in place. It also trails in the number of physical qubits in a working system, since charged atoms repel each other, and because every individual qubit has to be controlled by lasers. This makes charged-ion systems more difficult to scale.
NYSE: IONQ
Key Data Points
Two companies using this technology are IonQ (IONQ 2.10%) and Quantinuum (QNT +13.21%), both of which have achieved relatively impressive accuracy. IonQ has attained 99.99% 2-qubit gate fidelity, while Quantinuum has reached 99.92%.
The two companies are taking slightly different approaches, with IonQ turning to a combination of lasers and microwave antennas built on its chips to help shrink the size of its systems and improve stability. Quantinuum has stuck to using only lasers, arguing that using microwaves results in slower gate speeds.
3. The neutral-atom qubit approach
Like the trapped-ion approach, the neutral-atom approach also uses individual atoms suspended in a vacuum, but in this case, they are uncharged. They are controlled and manipulated using grids of tightly focused lasers, called optical tweezers. Since neutral atoms don't repel one another like ions do, it is possible to achieve a greater qubit density. This can lead to high accuracy and faster speeds.
Ultimately, the neutral-atom approach lies between the superconducting qubit and trapped-ion approaches. The technique is still thousands of times slower than superconducting qubits, while its 2-qubit gate fidelities trail those of trapped ions.
NYSE: INFQ
Key Data Points
While several pure-play quantum computing companies are pursuing this method, most are currently private, except for Infleqtion (INFQ 0.30%). Infleqtion achieved 99.73% 2-gate fidelity in 2024 and believes it can reach 99.9% this year. The company also has a strong quantum sensing and precision timing tools business. Alphabet is also exploring this technology.
4. Quantum annealing + superconducting
It's also worth mentioning D-Wave Quantum (QBTS 4.26%). While it is a leader in quantum annealing, that is more of a niche technology, as it's only suitable for a narrow range of uses, mostly involving optimization problems. That contrasts with the systems being built around the other technologies discussed above, which would be more general-purpose machines, useful in solving a wider variety of extremely complex problems.
However, through D-Wave's recent acquisition of Quantum Circuits, a company founded by a Yale professor who developed the superconducting circuit architecture, the company is also pursuing a unique hybrid technology that uses a dual-rail gate-model processor.
NYSE: QBTS
Key Data Points
The goal is to apply the quantum system expertise it has acquired while developing annealing systems to superconducting qubit technology to develop a system that combines the speed of superconducting qubits with the superior accuracy of trapped-ion technology. However, it's still early in that process, and the company has yet to provide any verified metrics.
