The computational power of a quantum computer arises from the superposition of quantum states, which gives a net parallel-processing size that exceeds the number of atoms in the universe for even a modest 300-qubit processor. To build such a computer, many groups around the world are exploring whether experimental systems can be controlled sufficiently well and with enough quantum coherence. A leading candidate in this search is based on “quantum integrated circuits” using superconducting devices. I will review current research that shows that the basic hardware requirements for a quantum computer can likely be met in the next few years, at which time a program to actually build a quantum computer can be started. An architecture based on surface codes allows one to estimate what a quantum computer might look like: even though the physical requirements are challenging—10 million qubits in a building-sized facility—it is much more realizable than an equivalently powerful classical computer, a continent-sized data farm.