[AndrewKemendo]

A few years ago I went to a seminar run by DWave and I really pressed them along similar lines that you were asking. A few of their engineers basically walked me through the concept.

As I understand it - and someone please correct me where this is wrong:

If you have designed an algorithm/program that optimizes for some minima (ex gradient descent), then you as the developer will need to map that algo into their system which basically creates a quantum topological map.

That topo is loaded into the QC and then it is run and outputs the "coordinates" of the minima.

Makes sense if you can 1. Convert your algo into something that can optimize for minima

2. Convert that into the topo space

3. Have no need for real-time results