Semiconductor spin qubits: Playing the long game

Today's noisy intermediate-scale quantum (NISQ) processors support of order 1000 qubits, yet resource estimates suggest that more than one million physical qubits will be required to achieve fault- tolerant quantum computation with the surface code [1]. Of all of the qubit technologies being pursued, semiconductor spin qubits most closely resemble conventional transistors, which can be fabricated at scale with ~100 billion transistors on a chip. It is therefore prudent to pursue long-game approaches to fault-tolerant quantum computing with semiconductor spin qubits. In this lecture, I will describe how the international research community has addressed some of the most pressing challenges facing silicon spin qubits (decoherence, large effective mass, and multiple conduction band valleys). Recent progress includes multi-qubit control with fidelities that are on par with competing technologies [2], as well as a variety of approaches for coupling modules of densely packed spin qubits [3,4,5]. I will also describe new device platforms that open the door to the fabrication of large two-dimensional spin qubit arrays [6].