At a time when most of the headlines around quantum computing are dominated by the big tech companies (IBM, Google, Microsoft), it is good to see the efforts of the IonQ company getting some attention. It is important to realise that there is more than one way to build a quantum computer and more than one measure to the quality of such a device. The big tech companies have specialised in building quantum computers using super-conducting effects, whereas as IonQ is blazing a different trail with "trapped ion" qubits.
Super-conducting qubits can take advantage of the optimised lithography techniques used in microelectronic design and have therefore been much more effective in increasing the number of qubits (which one can think of as analogous to the size of the registers and memory with which classical computers perform their operations). The number of qubits of a device is not the only measure of its power, however. One should also ask how stable the qubits are when they are asked to store or modify quantum information. By this measure trapped ion qubits perform better. Although, the flakiness of qubits can be ameliorated by error correction, this leads to overheads which increase greatly with the duration of the computation and trade-offs need to be examined closely.
Another important question is how well we can transfer quantum information between one technology and other (e.g. from super-conducting or ion trap to photons and vice-versa). This process known as transduction allows quantum computers to be built from hybrid technologies and for them to be networked for distributed processing. This could have a dramatic scaling effect on capabilities. Again, in this regard, ion traps are superior to super-conducting qubits.
The advantages and disadvantages of different qubit technologies make it wise to pursue research in all of them. We recall how early classical computers rapidly iterated through different technologies (electro-mechanical to valve to transistor to microcircuit) as the designs advanced. Thus, although the impressive "quantum supremacy" and "quantum advantage" milestones of the super-conducting devices are rightly celebrated, one should not disregard other approaches for future challenges.
IonQ are also admirable for focussing on the near-term applications of quantum computing rather than challenges from computer science. They actively engage with scientists involved in optimising chemical processes or designing pharmaceuticals to take advantage of physical simulations that can be done with quantum information, but are beyond classical devices.
Cybersecurity professionals are less interested in the simulation and sampling aspects of quantum computing than the ominous implications for certain forms of cryptography. The Internet is currently engaged in a large-scale migration to remove legacy forms of cryptography before a cryptanalytically relevant quantum computer becomes a threat. Critical to this effort is a sense of when such a computer might be available to various threat actors. A great deal of excellent work has been done by the likes of Craig Gidney to precisely estimate the quantum resources that could threaten certain legacy algorithms. However, these efforts have focussed very closely on the capabilities of the prominent super-conducting machines. A trapped ion device such as is being developed by IonQ would have different advantages and disadvantages to the super-conducting devices and the sweet spot for resource usage might be a quite different area than the current quotes.
Just as we should encourage the development of different technologies for quantum computation, we should also research how the distinct characteristics of technologies lead to different estimates for the resources required for an Internet-breaking quantum computer.
20 August 2025
Dr Daniel Shiu, Chief Cryptographer, Arqit