a question about num_read
Hello, I have a question regarding the use of num_reads. When one calls DWaveSampler.sample_qubo with numreads > 1, what is the schedule that the QPU uses for sampling the different reads? Is there a re-initialization of the qubits before proceeding to the next sampling or does the next sampling start from the point where the previous finished? Would this latter choice imply that two samples are somehow correlated? In practice, is there any difference in running DWaveSampler.sample_qubo once with numreads = 1000 and running it 10 times with num_reads=100?
Comments
Hello,
So, to answer your question about correlation between samplings, you can take a look at the following two links:
https://docs.dwavesys.com/docs/latest/c_qpu_errors.html#spin-bath-polarization-effect
https://docs.dwavesys.com/docs/latest/c_qpu_errors.html#temperature
Spin bath polarization effects can cause there to be some correlation between runs, and other jobs running on the chip could potentially cause temperature variations that have an effect.
As for the breakdown of anneal time and initialization time, there is a very thorough description of this in the following documentation:
https://docs.dwavesys.com/docs/latest/c_qpu_timing.html
I hope this answers your questions.
Please let us know if you need any clarification or more information!
Please sign in to leave a comment.