CMOS-OxRAM integrated circuits for low-power AI and cryogenic quantum dot auto-tuning

点击翻译按钮获取中文摘要

The latest major breakthrough in quantum computing (QC) has been the demonstration of quantum systems with more than 50 superconducting qubits allowing quantum supremacy for the first time. Other very promising qubit technologies include spin qubits based on Si, SiGe or III-V quantum dots (QDs). They leverage the great maturity of CMOS technologies and fabrication processes to offer low cost and highly scalable quantum systems at the chip level. Major research centers and industrials like CEA and Intel have started to report high quality spin qubits based on semiconductor technologies. However, the calibration and control of spin qubits are still performed mostly by hand with bulky classical electronics located outside the cryostat. The absence of fully integrated cryogenic electronics makes it impossible to build a large-scale quantum system due to the "wiring bottleneck" between the spin qubits and the control electronics. Additionally, QDs are prone to device-to-device variability inherent to manufacturing imperfections and environment perturbations. One of the next major breakthroughs in QC is thus to automate the control of very large numbers of spin qubits based on QDs using integrated in-situ cryogenic electronics.The partnership between 3IT and 1QBit aims to (i) demonstrate real time auto-tuning of solid-state QDs using low-power machine learning and In-Memory Computing paradigm based on the co-integration of CMOS circuits and emerging TiOx-based resistive memories (OxRAM) and (ii) design and fabricate ultra-low power artificial intelligence hardware specifically designed to operate at cryogenic temperature for in-situ tuning. This unique proof-of-concept will pave the way for large-scale quantum systems enabled by fast and energy efficient AI solutions co-located with quantum circuits inside a cryostat.