End-to-end workflow for machine learning-based qubit readout with QICK and hls4ml

TL;DR

This paper introduces an end-to-end workflow integrating neural networks into superconducting qubit readout hardware using QICK and hls4ml, achieving high fidelity and low latency.

Contribution

It presents a novel co-designed workflow embedding ML models into FPGA-based quantum readout hardware with quantization-aware training and efficient implementation.

Findings

Achieved 96% single-shot fidelity in qubit readout

Latency of 32 nanoseconds for the ML algorithm

Less than 16% FPGA resource utilization

Abstract

We present an end-to-end workflow for superconducting qubit readout that embeds co-designed Neural Networks (NNs) into the Quantum Instrumentation Control Kit (QICK). Capitalizing on the custom firmware and software of the QICK platform, which is built on Xilinx RFSoC FPGAs, we aim to leverage machine learning (ML) to address critical challenges in qubit readout accuracy and scalability. The workflow utilizes the hls4ml package and employs quantization-aware training to translate ML models into hardware-efficient FPGA implementations via user-friendly Python APIs. We experimentally demonstrate the design, optimization, and integration of an ML algorithm for single transmon qubit readout, achieving 96% single-shot fidelity with a latency of 32ns and less than 16% FPGA look-up table resource utilization. Our results offer the community an accessible workflow to advance ML-driven readout and adaptive control in quantum information processing applications.