Millions of qubits need to be employed in a quantum computer to achieve a fault-tolerant quantum operation. To reduce the complexity in such a large-scale system, the control and readout circuitries have been proposed to be placed at the 4 K stage of dilution refrigerators [1]. CMOS technology is commonly used to leverage its scaling to enable large integration of control and readout circuitries with qubits. However, the high-fidelity readout operations require low noise amplifiers (LNAs) with a noise temperature of a few Kelvins. This necessitates the usage of HEMT and parametric amplifiers [2]. Recently reported CMOS LNAs are still far away from attaining such good performance [3–5]. Thus, this is one of the greatest challenges on the way to the fully integrated CMOS readout. Additionally, due to the limited cooling power of dilution refrigerators, low-power solutions are needed for achieving a very good noise performance at 4 K. This paper presents a 28 nm CMOS LNA for qubit readout, which achieves an order of magnitude power reduction compared to its CMOS counterparts while still providing a similar good noise Figure (NF) performance at 4 K.
This letter presents an ultrawideband cryogenic low-noise amplifier (LNA) with current reuse and stacked inverter-based input-referred noise (IRN) cancellation in 40-nm CMOS. The first stage utilizes an IRN reduction high-
Qgate inductor with current reuse, a stacked inverter-based shunt-resistive feedback input stage, and a cryogenically aware back-gate biasing technique to regulate the threshold voltage under cryogenic operation. The second stage is the common-source (CS) main amplifier, followed by a cascode current-reuse auxiliary amplifier to suppress the channel thermal noise of the main amplifier. The third stage is the current-reuse CS inverter-based postamplifier buffer stage. At 4 K, the LNA achieves a measured peak gain (
S21) of 37 dB, with a large 3-dB bandwidth (BW) from 0.15 to 8 GHz and aNFminof 0.4 dB at 3.2 GHz under
PDCof 13.2 mW. To our knowledge, the proposed LNA achieves the BW bandwidth and the figure of merit (FoM) among the highest compared to published CMOS LNAs operating at 4 K. The proposed LNA is suitable for both spin and superconducting qubit readout.
Quantum error correction (QEC) is in demand for a fault-tolerant quantum computer. Currently, the QEC cycle is bottlenecked by the quantum bit (qubit) readout. To this end, RF dispersive reflectometry with frequency division multiplexing (FDM) is preferred due to its high fidelity and scalability. However, for a low noise temperature
Tnoise, cryogenic low noise amplifiers (cryo-LNAs) of classic LNA-first reflectometers require high transconductance efficiency (
gm2/Id). These open-loop cryo-LNAs also present limited linearity under large output voltage, resulting in errors of FDM readout. In this work, a new approach, a mixer-first cryogenic CMOS (Cryo-CMOS) reflectometer, is explored for its potential advantages in noise and linearity. A passive, low-loss mixer directly down-converts the RF reflection from the qubits (5.5–8 GHz) at 4 K to IF/baseband (5–250 MHz). Then, an IF/baseband amplifier, employing a MOSFET pair of$W/L = 2{\,}\times {\,}800~\mu $m/100 nm and current density of$5~\mu
A/\mu $m, achieves$2.8\times
higherg_{m}^{2}/I_{d}
thanCryo−CMOSLNAsforlownoise.Afterclosed−loopamplificationwithhighlinearity,thebasebandissampledbya500MSPS,12−bittime−interleaved(TI)analog−to−digitalconverter(ADC).Finally,thequbitstateisdiscriminatedbyanon−chipFDMdemodulationalgorithm.Fabricatedbya28−nmbulkCMOSprocess,themixer−firstreflectometerat4KachievesaminimumT_{noise} =19
Katf_{LO} =7.3
GHzwitha42−dBRXgain,whichishighlycompetitivetootherLNA−firstreflectometers.Inaddition,thankstotheco−optimizationofsingle−sideband(SSB)transmitter(TX),mixer−firstRXandon−chipdemodulation,underabasebandamplitudeofV_{pp}\sim 0.8$V, excellent spur-free dynamic range (SFDR) of SFDR > 45 dB and third-order intermodulation rejection of IMD
3>50dB are recorded. Lastly, the simultaneous FDM readout of two transmon qubits has also been demonstrated with fidelity 95.3% and 92.3% for qubit1 states$|0\rangle {\,}\& {\,}|1\rangle
,andfidelity96.1|0\rangle {\,}\& {\,}|1\rangle $, respectively.