What constitutes an "experiment" on LEAP?
On my free LEAP account, I see I have "4000 MAX EXPERIMENTS REMAINING", how does this work with reusing embeddings as discussed here? https://support.dwavesys.com/hc/en-us/community/posts/360016737274-Save-time-Reuse-your-embedding-when-possible What constitues an "experiment" for the purposes of that limit? Thanks!
Comments
Hello,
So the article you linked to is talking about something a little different.
The article is talking about classical computing time that happens locally on your computer.
It takes time to find an embedding, so using the same embedding will save you time when running the problem again.
The QPU time used on the other hand will be between 15 to 250 ms per problem.
Some time can be saved by doing multiple reads when submitting each problem (that is taking multiple samples on each problem submission).
This can be done by setting the num_reads parameter in the sample function.
This will save time, because the problem will not need to be programmed onto the QPU each time the problem is run.
Also, the size of the problem -how many variables, and the quadratic interactions between those variables- will determine how much time is used.
I hope that helped!
Please let us know if you have any additional questions.
Please sign in to leave a comment.