During the last five years, quantum computers have reached a breakneck pace of development. The most impressive milestone that we have witnessed so far came in October 2019: after several weeks of general uncertainty, the Google research team led by John Martinis published in Nature the scientific article in which he explained how he had achieved quantum supremacy.
This event made a deafening noise that lasted for months, but then came other milestones that confirm without leaving the slightest doubt that quantum computers are improving at a speed that a few years ago would have seemed unattainable. Without going any further, IBM, which is one of the companies that is bidding more decisively for this discipline, has already prepared Osprey, a 433-qubit quantum processor. And in 2023 he promises to have Condor ready, a quantum chip of no less than 1,121 qubits.
Intel, Google, Honeywell or IonQ are also dedicating a part of their resources to the development of this technology, but quantum computers, fortunately, are not only in the hands of governments and large companies. Many emerging companies have also entered this market fully, and some of them, such as China’s SpinQ or Australia’s Quantum Brilliance, have in their hands very promising innovations. Of course, all of them essentially pursue the same thing: to develop the first quantum computer capable of correcting its own errors.
There are several paths that lead us to fault-tolerant quantum computing.
During the conversation we had with him in June 2021, Ignacio Cirac, who is a Spanish physicist who heads the Theoretical Division of the Max Planck Institute for Quantum Optics and is unanimously considered one of the founding fathers of quantum computing, explained that developing computers quantum that do not make errors is very complicated. However, he also assured us that he is convinced that, despite the difficulties, these machines will arrive. According to him they will take time, but they will arrive.
Ignacio Cirac believes that in order to solve most of the problems that scientists hope to be able to tackle in the future with quantum computers capable of correcting their own errors, such as optimization errors or those in the field of cryptography or artificial intelligence, it will be necessary have several million qubits. Maybe even hundreds of millions of qubits. As we have seen, the most advanced quantum processor currently is IBM, and it has a few hundred qubits, so it is clear that there are many technological challenges that need to be solved.
Superconductors will probably help us get more qubits, but they are more prone to errors than ion trap qubits
The interesting thing is that there is no single way to go down this path. Organizations doing quantum computing research are working on a number of different qubit technologies, each at a different stage of development. IBM, Intel and Google are some of the big companies that have opted for superconducting qubits, but so have other much smaller ones, such as Atlantic Quantum, IQM, Anyon Systems, Rigetti Computing or Bleximo.
In fact, if we stick to the number of companies that are working on these types of quantum bits, it is reasonable to conclude that this is the technology that has the most support and investment, so, in a way, is the one that goes in the head. This strategy is probably the one that will help us to have more qubits, but it is also more prone to errors than ion trap qubits, which are one of the alternatives to superconductors. In addition, these last qubits are characterized by working at a temperature of about 20 millikelvins, which is approximately -273 degrees Celsius, with the purpose of operating with the highest degree of isolation from the environment possible.
As we have just seen, ion traps are currently the main alternative to superconducting qubits. This is the technology in which IonQ and Honeywell, among other companies, are working, and it is characterized by using ionized atoms, and, therefore, with a non-neutral global electrical charge. This property makes it possible to keep them isolated and confined inside an electromagnetic field, although this is only the starting point.
IonQ acts on the quantum state of its qubits by cooling them to reduce the level of computational noise and uses lasers to operate with them
From here IonQ acts on the quantum state of its qubits with ion traps cooling them to reduce the level of computational noise and uses lasers just below to operate with them. However, it does not use a single laser; it uses one for each ion, and also a global laser that acts on all of them simultaneously.
Honeywell also uses ionized atoms and lasers, but the procedure it uses to establish entanglement between two ions and act on them with a laser is different from that used by IonQ. In the cover image of this article we can see what Honeywell’s ion trap qubits look like, and they don’t look the least bit like superconducting qubits. It is logical that this is the case because, as we have seen, their technology is very different.
Ion-trapped qubits are more robust than superconducting qubits, allowing them to effectively dodge quantum decoherence for longer.
Superconducting qubits and ion-trapping qubits are currently the most developed, but other technologies are making good progress as well. One of them uses ions implanted in macromolecules, and is capable of storing information in them and performing simple calculations. In Spain there are several research groups working on quantum computing with molecules, although much work remains to be done in this field.
It is impossible to foresee which of these technologies will be able to deliver the number of qubits and the robustness necessary to develop an error-tolerant quantum computer.
Another very promising qubit technology is neutral atoms. It is also being pushed by various research groups, and is attractive because it is managing to collect many qubits while maintaining high accuracy and relative tolerance for errors. In fact, this strategy somehow combines the virtues of superconductors and ion traps, although much research remains to be done in this area as well.
Right now it is impossible to foresee which of these technologies will be able to deliver the number of qubits and the robustness necessary to develop an error-tolerant quantum computer. The technology that finally reaches this milestone may even be none of the four in which we have investigated in this article. Much remains to be solved, but a huge effort is being made to advance this discipline, and we can be sure that before the end of this decade we will witness enormous advances in quantum computing.
Images: Honeywell | IonQ