Superconducting quantum computation with imperfect resources

In this thesis, we address some important problems in current research in super-conducting quantum computation. We first describe how a quantum computer can be constructed using the rf-SQUID as the qubit, which was first proposed by us. We explain how the “macroscopic quantum coherence” associated with the rf-SQUID can be used to realize a qubit. We demonstrate the possibility of realizing a universal quantum computer with our superconducting technology, by showing explicitly how single bit and two bit operations can be performed. We then go ahead to address one of the most serious problems in the way to the construction of a practical superconducting quantum computer, which is the lack of variable couplings between qubits. This problem is serious in the sense that it is seemingly intrinsic to superconducting quantum computing, because the value of the couplings is fixed by the fabrication and cannot be varied during the computation. After reviewing previous unconvincing solutions to this problem, we discuss in detail our unique solution, which relies on the idea of computing with encoded qubits. We show that if we compute in and out of carefully designed “interaction free subspaces,” we can effectively switch on and off couplings between logical bits, therefore realize universal quantum computation without pulling any physical switches between qubits. Our unique approach allows us to build a fully scalable superconducting quantum computer using current available technology, with only a modest overhead in the number of qubits. Having solved this problem, we then discuss experimental techniques that would allow us to study the macroscopic quantum coherence. We show how we can demonstrate macroscopic quantum coherence by observing the Rabi oscillations in the rf-SQUID, taking advantage of a technique to effectively switch on and off the microwave quickly. This experimental method is much less demanding than our previous design of the experiment, therefore it is more likely to succeed in a short period of time. Circuits that implement these ideas have been laid out and are ready to be fabricated.