Putting D'Wave's 'Orion' to the Test - (Part 4)

....Continued from Part 3

Note: I've heard (today: 26 August 2009) from the folks at Orion web services and they told me that my optimization job failed because it exceeded the maximum quota...for the job run time (10 mins). Looks like I'll have to figure out news ways of streamlining my problems!

As my problem was successfully 'parsing' through Orion, I was thoroughly enthralled by the thought of parallel universes entangling, at that very moment, to wrestle the binary quadratic portfolio optimization problem that I had submitted. Unfortunately, my extreme excitement was destined to morph into an admixture of disappointment and anger.

Approximately ten minutes after I had submitted my problem into the Orion application, I received this message from the application:

Click on Image for a Better View

Hence, I attempted to figure out what went wrong; was there an error in the problem structure, or in the parameters that I set when I submitted the problem?

I revised my problem matrix, and the parameters that I had set, and everything appeared to be sound:

Click on Image for a Better View

This part of the Job status revealed what went wrong:

Click on Image for a Better View

Apparently, Orion's Binary Quadratic Optimization Annealing Solver had completed the process of computing my portfolio optimization problem, but it timed-out just before it could produce the solution... Strange.

Maybe there could be a bug in Orion, or I have just witnessed the effects of decoherence in a quantum system (after all, quantum data is very fragile). I know that Orion is currently powered by an interconnected computing grid that simulates a quantum computer. Maybe something perturbed the grid whilst it was compiling the answer.

I'll try again at random intervals just to make sure.