D-Wave Systems, located in British Columbia, is a builder of commercial quantum computers. It stores bits as magnetic directions in one of three states: clockwise, counterclockwise, and both directions simultaneously. The math and physics are far beyond me, but they claim to solve certain sets of optimization problems up to 100,000,000 times faster than classical computers. Customers for their computers, which cost $10 million apiece, include Lockheed Martin, an unnamed intelligence agency (NSA?), Google, JPL and NASA Ames Research.
Applications appear to be computationally intensive problems with lots of variables, and the solution involves a process called quantum annealing, where an optimal approach is found by exploring millions of solutions simultaneously to find the most efficient solution path. I’m reminded of a discussion on the famous double slit experiment, a classic physics experiment that demonstrates photons displaying behaviors of both waves and particles, known as wave-particle duality. Most interesting is that quantum probabilistic behaviors are also observed, in that the experiment functions differently when the particle paths are observed and when they are not. When the photons in the experiment are observed, the probability function collapses and the photons behave like a particles. If they are not observed, the photons take many paths through the slits and create a dispersed pattern on the target. That behavior has been described as “spooky”, because the particles seem to know when they are being observed. Weird, I know. It’s been said that anyone who claims to understand quantum mechanics is lying. But that doesn’t mean we can’t describe its behavior. Richard Feynman explained that at the quantum level, every possible path a photon can take is considered, and the path chosen is a probability function, like a bell curve. As photons are emitted from a source, the most likely path is taken most often, but some photons will take slightly less probable paths, still other even less probable paths, and so on. Quantum annealing seems to be a form of that, where many paths are simultaneously considered until a most probable path emerges, then it is chosen.
What this means, I’m not sure. I think it means more than faster gadgets. The NSA is no doubt applying this to encryption and decryption. NASA AMES and others are applying this to artificial intelligence research. Lockheed could be applying this to problems as diverse as stealth shaping and calculating the highest probable intercept solution of a target in a ballistic missile defense engagement.
At this point, these machines require a low vibration, supercooled, electromagnetic interference free environment. Not something you’re going to put on a ship or spacecraft. A bit like early clocks in that sense. A clearly powerful technology in its infancy. But also clearly a technology with immense implications. What could be done with a technology that brings computing power hundreds of millions times greater than we currently have to your fingertips? How would that change things? How has the world changed since ENIAC was built? It’s an interesting development and I’m glad to see things like this still being built. It’s a good sign for our civilization.
5 thoughts on “Catch D’Wave”
I remember a classic problem from the days I was programming an ancient IBM 650.
We used to talk about a problem of calculating paths in a maze. I’ve forgotten the details (It was 57 years ago) but the story was that the computer could calculate the shortest path through the maze if it could stop once it had reached a sum greater than the previous largest number but, if it was required to calculate the length of all paths before concluding which was longest, it would take a million years or some number like that.
I expect these machines, or even my laptop, could do it in seconds. When I was doing masters at Dartmouth, I had an IBM PC and an old Mac PowerBook 150. We were doing regression analysis with multiple variables. One problem I was working on had 8 variables. The PowerBook would take 15 minutes to do the calculation. The PC would do it in a second. There was that much difference between them then.
Some scientists believe that quantum entanglement suggests the possiblity of God acting on the physical world
These results have intriguing philosophical implications, he notes, especially for the spiritually inclined. “You could say the experiment shows that space-time does not contain all the intelligent entities acting in the world because something outside of time is coordinating the photons’ results,” Suarez says. “Physics experiments cannot demonstrate the existence of God, but this test shows that today’s physics is compatible with all major religious traditions. There is strong experimental evidence for accepting that nonmaterial beings act in the world.”
I am not a physicist, nor am a quantum mechanic*, but, if quantum phenomena are probablity functions, how do we know that a quantum computer is giving us the right answer to a question? Or is just going to give us odds, like Vegas?
*A joke from the old Car Talk show.
I do not know Robert. But taking a stab, it seems the bits are encoded as ones and zeros like in classical computers. However, the approach to the solution is found through probabilistic analysis.
An analogy I heard was a mapping problem. Suppose you wanted to find the path of least changes in elevation through a hilly region so that a road could be put in most efficiently. You could walk the entire region taking measurements. Quantum annealing allows all those paths to be explored simultaneously, like Feynman’s photons (above) considering all paths to a target before collapsing the probability function to an actual path. Once the optimal path is determined, then the road can be built. Once the optimal path to the solution of a 1000 variable problem is found, that path is then used to do the calculation. The hard part was figuring out the best means of approaching it. I think that’s basically it. I could be completely wrong.
“Quantum annealing allows all those paths to be explored simultaneously, like Feynman’s photons (above) considering all paths to a target before collapsing the probability function to an actual path. Once the optimal path is determined, then the road can be built.”
Yes, but, from what you are saying, the odds are that you will find the shortest path. OTOH, sometimes the long shot comes in. A lot of folks bet a lot of money on Cam Newton and the Panthers, but Old Man Manning (who is going to retire tomorrow) and the Broncos won the game.
A lot of life is playing the odds, and it is a good thing to know what they are, but it is not the same thing as being sure of the right answer.
Comments are closed.