Energy-Efficient and Robust Cryo-CMOS Smart Temperature Sensor

Abstract

Quantum computers, mainly rely on quantum bits (qubits), which need to be operated at cryogenic temperatures. With the need for systems with thousands or even millions of qubits, the challenges of wiring and scalability become more apparent. To address this, several studies focused on the idea of fully integrated cryogenic controller systems. It operates in close proximity to qubits in a deep cryogenic environment. In order to ensure operation over a specific temperature range, accurate temperature monitoring by temperature sensors (TS) is critical in this type of setup. Such sensors should operate efficiently at low temperatures and should operate over an extremely wide temperature range (from 4K to 300K) with high accuracy and energy efficiency. Current TS still faces performance challenges in such applications, especially in temperatures below 200K. CMOS-based smart temperature sensors have the advantages including low cost, compact size, and ease of use. Among the various CMOS sensing elements, Capacitively Biased Diodes (CB-Ds) were chosen for this project. Therefore, in this project, a CB-D-based temperature sensor with a novel hybrid voltage-time domain readout in Intel 16nm FinFET technology is proposed with the capability of operating within a range from 4.2K to 300K.

Based on this topology, a further over-ranging technique is employed, thereby energy efficiency and accuracy are improved. From simulated data, with the supply VDD = 0.90V, an accuracy of +0.5/ − 0.3K with a conversion time of 44.1 µs is achieved. Furthermore, it consumes 21.9 µW, which has a competitive performance with room temperature prior art smart temperature sensors. Based on measurements, a customized smart temperature sensor can be designed. This will allow cryo-CMOS thermal monitoring system to be designed, which will provide an important milestone in the realization of scalable quantum computing.