| Summary: | We are rapidly entering the era of potentially useful quantum computation. To keep on designing larger and more capable quantum computers, some form of algorithmic noise management will be necessary. In this thesis, I propose multiple practical advances in quantum error mitigation and error correction. First, I present a novel and intuitive way to mitigate errors using a strategy that assumes no or very minimal knowledge about the nature of errors. This strategy can deal with most complex noise profiles, including those that describe severe correlated errors. Second, I present proof that quantum computation is scalable on a defective planar array of qubits. This result is based on a two-dimensional surface code architecture for which I showed that a finite rate of fabrication defects is not a fundamental obstacle to maintaining a non-zero error-rate threshold. The same conclusions are supported by extensive numerical studies. Finally, I give a new perspective on how to view and construct quantum error-correcting codes tailored for modular architectures. Following a given recipe, one can design codes that are compatible with the qubit connectivity demanded by the architecture. In addition, I present several product code constructions, some of which correspond to the latest developments in quantum LDPC code design. These and other practical advancements in quantum error mitigation and error correction will be crucial in guiding the design of emerging quantum computers.
|
|---|