Context
Quantum computation promises to speed–up certain problems of interest that would be infeasible to solve using today’s (classical) computing paradigms. Potential applications include quantum chemistry, material science and machine learning. Although much progress towards the implementation of quantum computers has been made in the past 10 years, there remain formidable challenges. A significant problem is posed by the debilitating effects of noise which is inevitably present in quantum devices. It is widely accepted that, unlike for classical computers, it is not possible to increase the reliability significantly by better engineered systems. Thus, in order to deal with noise, a whole toolkit has been developed. A central part of this toolkit consists of quantum error correcting codes which are analogous to error correcting codes used in data storage devices and wireless communication. While in principle we now know how to correct errors in quantum devices, it has been shown that the amount of resource overhead is still forbiddingly high. The resource overhead for fault-tolerance is problematic, as the number of physical qubits will be limited in the early stages of building quantum computers. Further, certain no-go results show that quantum information that is protected by a quantum code is hard to manipulate. These no-go results can be circumvented by expending more physical qubits, putting further strain on resources. This poses a significant obstacle for the viability of quantum computing in the near- and mid-term.
Objectives
The overall objectives for the associated team is to establish strong foundations for fault-tolerant quantum computation in realistic and practical settings. This will participate to the global effort of bringing to life reliable quantum computation as soon as possible. For this, a systematic understanding of the fault-tolerant computation capabilities of small quantum codes is required. This includes small block codes using qubits as well as small codes leveraging hardware control for error correction like bosonic codes. These kind of small setups are the most mature today. In the near future it will be possible to link them in a modular way and that is why anticipating distributed architecture is also a priority.
