Cryogenic CMOS Compact Modeling for Cryo-Electronic Applications

Quantum computing holds the promise of a monumental leap in computational power, enabling the resolution of previously insurmountable problems with astonishing speed compared to classical computers. \nEmerging computing paradigms, including Shor's factoring algorithm, Grover's searching algorithm, quantum simulations, protein folding, and more, stand on the brink of feasibility, thanks to quantum computers. \nHowever, despite impressive recent advancements in quantum computing, demonstrated systems featuring anywhere between a few to around a hundred physical qubits remain significantly distant from achieving quantum supremacy over classical computing, which demands the utilization of millions of physical qubits. \nThis formidable challenge is known as the scaling problem. \n \nImplementing large-scale quantum computing systems is faced with numerous hurdles, particularly \nwhere each qubit necessitates precise control under extremely low cryogenic temperatures (<1 K). \nComplementary metal-oxide-semiconductor (CMOS) technology, the cornerstone of classical computer scaling, emerges as a promising solution for scaling quantum computers. \nCMOS technology offers deep miniaturization and versatility, functioning seamlessly at both room temperature (RT) and cryogenic temperatures (cryoT). \nCMOS is compatible with the spin qubits in semiconductor quantum dots (one of the various methods of implementing qubits that exhibit long coherence time) offering integration compatibility especially from the fabrication perspective. \nIt is this kind of tight integration that may ultimately hold the key to resolving the quantum scaling problem, bridging the gap between the current state of quantum computing and its promising potential. \n \nNevertheless, current circuit design environments lack support for operating temperatures \nnear cryoT. \nThis lack of support is centered in the often overlooked component known as the compact model. \nCompact models act as the blueprint that informs circuit simulators of how circuit elements behave under various operating conditions. \nThis component is composed of simplified mathematical formulas that bridge the gap between the element's physical model and simulation engines. \nIn order to obtain accurate simulation results necessary for cryo-circuits design the compact model must be accurate. \nTo obtain precise simulation results necessary for cryo-circuit design, it is imperative to understand and incorporate the effects of cryoT on metal-oxide-semiconductor field-effect-transistors (MOSFETs) into the compact model. \n \nThis thesis is one of the first attempts to develop cryogenic MOSFET compact model based on virtual source concept, through theoretical investigation and experimental validation. \nThe cold temperature effects on MOSFETs are studied and integrated into the existing MIT virtual-source model (MVS), expanding its temperature range to include deep cryoT in the range of few Kelvin. \nTo achieve this, sample devices from multiple commercial technology nodes are characterized in RT down to deep cryoT. \nThe thesis outlines the measurement setup and explores a range of predicted and unexpected cryogenic phenomena within the transistor. \nGood agreement between experimental data and modeled data is obtained between 300 K and 4 K for 20, 28, and 65 nm bulk CMOS technology nodes. \n \nHowever, merely developing a cryogenic compact model is insufficient for its adoption and practical deployment. \nExtraction and fitting tools are therefore developed alongside the model. \nTo support circuit design on industrial tools and validate the model, the compact model is implemented Verilog-A. \nSubsequently, a cryogenic circuit simulation is demonstrated using industry-standard electronic design automation (EDA) tools. \nThis demonstration underscores the viability of the model to facilitate cryo-circuit design for quantum computing, representing a significant step towards realizing the potential of quantum computing in practical applications.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.