With Yosi Rinott and Tomer Shoham we studied various statistical aspects of samples coming from NISQ computers.
My question is about available data consisting of samples from NISQ computers
We would be very happy to get samples from runs of quantum circuits (along with the description of the circuits). If such data is publicly available this will be especially useful.
Here are some specifics about what we would be happy to get and what we intend to try.
We will be happy to receive samples of 0-1 strings coming from NISQ circuit. Circuits with n qubits where n is between 5 and 30 will be very helpful for us. It will be helpful to have sufficiently many samples so we can get a good picture of the empirical distribution.
What kind of circuits? We will be happy to get any kind of circuit along with the description of the circuit itself. Especially useful for us will be random circuits similar to those that appeared in the Google 2019 "supremacy" experiment, or other kinds of random circuits.
Calibration: If there is a preliminary process of calibration (similar to Google's) so that one haa a description of circuits based on "standard" gates and then a better description of the same circuit based on adjusting the gates to the device, we will be happy to have both descriptions of the circuits.
What we plan to do with the data:
a) For random circuits (and perhaps also for other types) we will try to compute various estimators for the fidelity
b) We plan to compare the empirical distribution with theoretical distribution under various noise models
c) We plan to study (non)stationary and chaotic phenomenon in the empirical data
We are especially interested from data obtained by the IBM quantum computers. The IBM researchers expressed interest in random circuit samples many years ago, but we are not aware of experiments the IBM team or people working with IBM quantum computers conducted in this direction. Any links for data from IBM QC will be most helpful.
Data from simulators of noisy circuits
- We would also be happy to get samples based on simulations of noisy NISQ computers. Also here any reference to publicly available data will be helpful.