Casey Johnston at Ars has an excellent write up of a promising quantum computing experiment. While the efficiency of quantum computers at crunching certain classes of algorithms that are expensive on classical systems is still open very much to debate, another area I don’t see as often discussed is their use for the simulation of physical systems. Given the quantum oddities involved in building such simulations from first principles, the case for doing so on a quantum computer more easily than a classical one seems more plausible.
What Casey describes is a pretty simple optical setup with what I am guessing is a custom logic gate, not a general one like a c-not. Error introduced by the scale of the system is the single biggest limiter in applying it to larger molecules. I’d be curious to know, if we assumed that error was not an issue, that it was a solvable problem, what the energy and cost would be like for such a rig as we increased its scale up to everyday objects. Such an exercise could potentially validate Rudy Rucker’s contention that there lies a huge gap between a physical system and any computational simulation, no matter how efficient, cost-wise.
I will also admit to being curious about whether any form of this system could be applied to better understanding the exotic particle physics resulting from particle colliders like the LHC. I suspect that if there were, considering the primitive state of this one experiment, they are probably years off.