Google’s director of engineering, Hartmut Neven, has announced the strides achieved by Google’s Quantum artificial intelligence team, noting that the progress made can now be applied to fully understanding quantum annealers.
Don't Miss: The Best HDR TVs
The knowledge already acquired over the past couple of years the team has been working on comprehending the fundamental physics surrounding quantum annealers has been used to develop proof-of-principle optimization issues, while injecting these to the D-Wave 2X quantum annealer project that Google is carrying out with NASA.
The issues surrounding the study was made to show that the field of quantum annealing can possibly provide runtime benefits that could be applied to hard optimization issues that are based on tough energy terrains.
The engineers discovered that quantum annealing is much better than simulated annealing operating on single core when it comes to solving problems with about 1000 binary variables. When the researchers analyzed the results of quantum system with the Quantum Monte Carlo algorithm running on normal processors, they found a varying factor that is as large as 108.
Although much has been gained theoretically with quantum annealing, converting these successes into practical technology is where the work lies; and to this extent, future generation annealers that would be developed must be designed to incorporate concerns that have practical applicability. One of these problems might be enlarging the density and precision of control for connections existing among qubits together with how coherent they are.
To this end, Google’s quantum hardware group is trying at the moment to make it possible for users to enter hard optimization problems that would be solved by the system; and where higher order optimization issues arise, rugged energy terrains might be normalized to enhance finding solutions to the problems.
Don't Miss: See the first leaked Black Friday 2016 Ad
The team understands for now that quantum annealing offers better alternatives than simulated annealing, and they are positive that the considerable runtime gains they have obtained in the study can be converted to commercial use in daily tasks that require quality machine intelligence.