We announce the launch of a revised AI-TWPA-C equipped with an on-chip pump coupler that provides advances on both fronts. The path towards quantum utility is paved with hard-earned hardware and software advances. Among the hardware challenges superconducting quantum computers face when scaling up, one is related to qubit readout: how does one deal with the increasing number of qubit readout signals in the limited volume and mass budget available in dilution refrigerators? To address the challenge, one can frequency multiplex more qubits per readout line and/or decrease the volume taken up by each readout line.

We are eager to see what creative ways system integrators and researchers will come up with to push the boundaries of readout with this expanded frequency band. In more practical terms, the 5 GHz bandwidth also makes it easier to pivot to a different readout frequency regime, if needed. This is what we at Arctic Instruments have been focusing on: simplifying the deployment of near-quantum-limited readout chains.   

Let’s look at the impact of the product revision from a practical engineering perspective and to understand how adopting our technology will improve the day-to-day life of your business or research group. In particular, let’s compare the wiring diagram for our revised AI-TWPA-C with the conventional solution that uses an external coupler: 

With diplexer (or directional coupler)

With on-chip pump coupler

Reduced component count and wiring: higher quantum efficiency and fewer mistakes   

With the pump coupler integrated on chip, you avoid at least one external component (diplexer or directional coupler), one coaxial cable, and one SMA-SMA connector interface. This means that you save 0.5 to 1.0 dB of insertion loss in the most critical part of the wiring, before the first amplifier, which directly translates to improved quantum efficiency of the readout system by a factor of 1.12 to 1.26. Furthermore, there are fewer possible points of failure, in particular fewer possibilities for reflections caused by worn out or improperly mated connectors, which have a direct impact on TWPA performance.

Reduced component count and wiring: less weight and volume  

We’re looking at a decrease of roughly 100 g in weight on the mixing chamber plate per readout line with the simplified chain, and a corresponding decrease in volume. This means less material to cool and more space for your engineers to use for other things.  

Pump coupler at output: increased robustness and reduced pump power to the rest of the amplification chain  

Our revised TWPA also includes an on-chip-integrated pump coupler at the output that guides power to a termination resistor. Besides reducing concerns with pump-power-induced saturation in later amplification stages, the integrated coupler at the output makes the performance of the TWPA itself less dependent on the quality of impedance matching provided by the components placed at its output.  

Wider bandwidth and higher upper frequency: multiplexing more qubits and exploring higher readout frequencies

There has been a trend in recent years in transmon-based QPUs to move to higher readout frequencies. One clear benefit in doing that is that it is easier to achieve a lower thermal photon population in readout resonators when their frequencies are higher. With our increased bandwidth, it is possible to continue this trend up to 9.5 GHz. Our 5 GHz of instantaneous bandwidth also makes it easier to multiplex even more qubits per TWPA.    

Our revised AI-TWPA-C provides simplicity and improved performance that ultimately turn into time savings for the day-to-day work of scientists and engineers tasked with setting up and maintaining the readout lines. Together with the high level of stability and robust setup procedure of our amplifiers, AI-TWPA-C offers a remarkably low cost of ownership over the lifespan of the product.  

If you want a quote or simply want to know more about our product, contact us at info@arcticinst.io.