Abstract: Quantum computing hardware has made rapid strides in both qubit count and error rates since the first system was put online in 2016. Yet despite these advances, every quantum computing device available today is still susceptible to noise that ultimately limits its computational reach. Fortunately, it is possible to mitigate errors, on average, if one pays a price in terms of an exponential increase in the number of samples needed to reach a target precision. Utilizing mitigation techniques, it may be possible to perform useful quantum computation before reaching fault-tolerance.
In this talk we will look at the mitigation of measurement errors on quantum processors. Measurements have the largest single-operation error rates on platforms such as superconducting processors, and removing this source of errors yields a sizeable increase in fidelity for many near-term algorithms and applications. We will present a matrix-free measurement mitigation method (PRX Quantum 2, 040236) called M3 that, unlike earlier techniques, is scalable beyond a handful of qubits. Our method accommodates both uncorrelated and correlated errors and allows for the computation of accurate error bounds. We will demonstrate the validity of our procedure by showing that it is possible to mitigate errors in a few seconds that would otherwise be outside of reach for even the largest of supercomputers.
Hosted by PandA Colloquium Committee