dwavesystems
What happens when the number of variables doesn't fit on the Chimera topology
I am trying to solve problems with many variables using DWaveSampler for the Chimera topology. The sampler offers a solution as long as the number of variables doesn't exceed the number of qubits (2048). However, when we try to solve for example a complete graph of 200 variables, there is no embedding that fits the chimera topology, so some kind of descomposition in parts of the problem has to be made.
I know that when you call the LeapHybridSampler, a descomposition with classical algorithms is made, but I couldn't see which are the exact algorithms, as the code is closed-source.
My question is: Are the quantum-classical algorithms used to solve the problem using DWaveSampler the same as the ones used in LeapHybridSolver when we have many variables?
Thanks in advance
Hi @pablote1997 , unfortunately the specific way that the LeapHybirdSampler works is, as you say, closed-source. If you're interested in seeing how some hybrid algorithms handle solving problems with many variables, I would encourage you to check out dwave-hybrid.
To add to @arcondello's link, I would suggest also the resources of https://github.com/dwave-examples/hybrid-computing-notebook
