search

dwavesystems / qbsolv

Qbsolv,a decomposing solver, finds a minimum value of a large quadratic unconstrained binary optimization (QUBO) problem by splitting it into pieces solved either via a D-Wave system or a classical tabu solver. (Note that qbsolv by default uses its internal classical solver. Access to a D-Wave system must be arranged separately.)
https://docs.ocean.dwavesys.com/projects/qbsolv
Apache License 2.0
914 stars 229 forks source link

qbsolv and the signed particle formulation of quantum mechanics #24

Open jmsellier opened 7 years ago

jmsellier commented 7 years ago

Hello,

I am an Associate Professor at the Bulgarian Academy of Sciences and the creator/developer/maintainer of several GNU packages which can be found here: https://www.gnu.org/software/archimedes/ http://www.nano-archimedes.com/ https://www.gnu.org/software/gneuralnetwork/

In particular, nano-archimedes is a simulator for the quantum transport of electrons based on the signed particle formulation of quantum mechanics (which I came up with a few years ago). A paper on this formulation can be found here: http://www.sciencedirect.com/science/article/pii/S0021999115003708

It turned out that this formulation is computationally very advantageous (compared to other formulations). As a matter of fact I have been able, for example, to use it to simulate systems of indistinguishable Fermions on relatively small computing systems (actually a single-CPU computer): http://www.sciencedirect.com/science/article/pii/S0021999114006627

I also used this formulation to depict a quantum system exploiting interacting electrons which can solve a system of linear equations reformulated in terms of an optimization problem: http://link.springer.com/chapter/10.1007/978-3-319-21133-6_3

Many other papers can be found here: http://www.nano-archimedes.com/publications.php

Since, as far as I understood, the flux qbits used on D-WAVE systems can be considered as one-dimensional devices, it would be useful/interesting to solve the sub-QUBO problems by means of a simulated system by means of the signed particle formulation. For example, for relatively small sub-QUBO problems, one could embed nano-archimedes into qbsolv and solve these problems by means of a simulated set of one-dimensional electrons. This could bring to a more realistic simulation of the whole optimiziation process. What do you guys think?

I hope, somehow, this suggestion helps. I look forward to hear your opinion.

Thanks,

JM Sellier

jmsellier commented 7 years ago

Hi everyone,

I was wondering if anyone has something to say on the comment above. I personally think it is not a bad idea and I would like to give it a try. But before I do anything, I really would like to know what you D-wave guys think about it.

Thanks!

JM