Large scale quantum processors require high quality qubits and good inter-qubit couplings. Superconducting circuit is one of the most promising platforms for building NISQ era devices and executing error correction protocols due to its high two-qubit gate fidelities and clear path for scaling up. Despite the fast development on superconducting circuits (mostly based on transmon qubits), the two-qubit gate fidelities still need two orders of magnitude improvement to realize quantum error correction for solving real-world problems. The fluxonium qubit, which this thesis will be focusing on, is a very promising candidate due to its long coherence times, large anharmonicity and large parameter space for design diversity. By using low frequency fluxonium qubits, we achieve longer coherence times despite having the same material limitations. In this thesis, we demonstrate fast flux pulses that can finish arbitrary single qubit gates within a qubit Larmor period, while not introducing any DC components to the flux line. We also present the two-qubit gate protocol that works with this scheme, which only takes a few Larmor periods for a $\siswap$ or $\sbswap$ gate. We will discuss how the low frequency qubits can have long coherence times and how to operate the qubit despite the thermal heating. We believe that there is a path for fluxoniums to gain a large factor in performance compared to the current transmon systems based on the improved control designs.