Qubits are the heartbeat of quantum computers. Their hard-to-imagine properties are what gives quantum machines their awesome computational power.
Superconducting devices, spinning atoms, polarized photons, quantum dots, and trapped ions are not futuristic video games. They are different qubit technologies. Moreover, each qubit type has its peculiar advantages and disadvantages.
Out of all the qubit types, superconducting is the most common. However, trapped ion qubits, a relatively new qubit technology, shows a great deal of promise. In addition to having faster gate speeds, superconducting qubits are solid-state fabrications. On the other hand, trapped ions are more stable and have better connectivity to other qubits than their superconducting counterpart.
Functionally, all qubits depend on strange quantum properties. Instead of classical bits - a one or a zero - a quantum computer’s qubits (quantum bits) can be coded as a one, or a zero, or both a one and a zero. Qubits can also exist in all the possible states at the same time. That condition is called superpositioning.
Bits in a classic computer act individually, while quantum properties allow qubits to become "entangled" with each other. Once entangled, a group of qubits can act as a single qubit. That enables a solution to multiple inputs to appear on a single qubit.
Compared to classical computers that can only work on one computation at a time, superpositioning gives quantum computers the potential to execute millions of simultaneous operations.
Quantum teleportation is another quantum feature. It sounds like science fiction, but it's not. Instead of teleporting matter, quantum teleportation is limited to sharing quantum states between entangled particles regardless of how far apart they are. In the future, teleportation will be useful for controlling and using qubits in remote quantum servers as well as for telecommunications.
Superconducting heavy hitters
It's worth noting that Intel is not dependent on superconducting qubits. It is also investigating another qubit technology that operates in silicon, called spin qubit. The quantum state of spin qubits depends on the spin of an electron on silicon. One reason for Intel's interest in spin qubits is because it is another technology that can leverage Intel's vast experience in silicon manufacturing.
Rigetti Computing, a recent but impressive California start-up, also uses superconducting qubits. It is a full-stack company that beefed up its application development capability by a recent acquisition of QxBranch.
Superconducting qubit quirks
Superconducting qubits are the most mature of all the qubit technologies. That means we know what improvements are required even though we may not yet know how to do it. Superconducting qubits also have the advantage of being built using existing semiconductor techniques.
Superconducting qubits have a few disadvantages:
- They require near absolute zero temperatures to operate
- They are very susceptible to quantum noise
- They retain their quantum states for short periods
- Limited gate connectivity to qubits
IonQ, an upstart ion startup
IonQ, Alpine Quantum Technologies (Austria), and Honeywell all use trapped ion technology. However, IonQ is its driving force. IonQ was founded in 2015 by Christopher Monroe and Jungsang Kim. Monroe is the Bice Zorn Professor and a Distinguished Professor of Physics at the University of Maryland and Fellow of the Joint Quantum Institute. He is currently the Chief Scientist for IonQ. Kim is a professor in the department of electrical and computer engineering at Duke University.
Building a better ion trap
Trapped ion technology isn’t a radically new concept. It’s used to make some of the most accurate atomic clocks in the world.
Like atomic clocks, IonQ uses an isotope of ytterbium to build its qubits. They start with a neutral atom of ytterbium, then use lasers to remove an electron from the atom’s outer shell. This process converts a regular atom of ytterbium into a ytterbium ion (Yb+).
The ytterbium ion is held in place by electromagnetic fields in a linear ion trap. According to IonQ, because this technology is easy to reconfigure, they can load a hundred or more ions in a linear chain. Also, they can do it without the need to fabricate a new chip. So far, they have used single-qubit gates on a linear chain of 79 ions.
There are many advantages to trapped ion qubits. Compared to superconducting qubits, they need less overhead for error correction. Entangling groups of qubits in a shared trap is easy due to the Coulomb force. Another big plus is the fact that dilution refrigerators are not needed.
Long term view
There is much work to be done before a universal fault-tolerant quantum computer is available. The long-term viability of superconducting and trapped ion qubits looks good. Superconducting qubits will steadily improve as a result of the financial resources of our biggest and best tech companies.
If trapped ion computer researchers can solve the scaling problem with lasers, they have a good chance of exceeding the capabilities of their superconducting counterparts.
Almost all researchers agree we are in the early experimental stages of quantum computing. Best estimates are it will take another 15-20 years for quantum computing to reach maturity. There is an excellent chance that future research will discover better qubit technologies or materials.