Does the simulator of the IBM-Q supercomputer involves a CNOT gate error rate ?

What are the main differences reasons in outcome between simulators and real backends in IBM-Q ?



The IBM Q simulator does not incorporate any error rates. The simulator is set to run in an "ideal" environment, so you do not see the affect of any noise or errors. It is possible to add a noise model to your simulation to have the run be closer to running on the real device, but by default this is not set.

The main difference between the outcomes from the simulator and the outcomes from the real devices are the effects of noise. The simulator will only ever give you your expected results, whereas the real devices will return some outcomes that are unexpected. The real devices are affected by gate errors, readout errors, general environment noise, and decoherence and relaxation of the qubits. All of these things contribute to the few unexpected outcomes from the real devices.

  • $\begingroup$ First you say The main difference between simulator outcome and real backend outcome is the effect of noise, and then you say it's a set of reasons ( the cited 4 ). Is it the 4 reasons altogether that make the difference in outcomes from simulator and rea backend ? $\endgroup$ – user1319236 Aug 1 '19 at 20:04
  • 1
    $\begingroup$ The general reason is the fact that the real devices introduce noise. Some of the reasons I listed are components of noise. The gate and readout errors exist because of the noise created when applying gates and measuring respectively. Granted, decoherence and relaxation are more related to the qubit itself, but I felt the need to list them as they are still very important factors to remember when using the real devices. You are right though, I could have worded my answer a little better. $\endgroup$ – Matthew Stypulkoski Aug 1 '19 at 20:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.