Quantum computer leap | Hacker News Reader

Quantum computer leap

phys.org

24 pointsmikecane8 comments

Sorry, but no this is not a quantum computer leap. It's a cute paper but woefully far from a complete solution.

Some problems 1) the procedure appears to require 100 percent detection of photons (there is no analysis of this in the paper). This is not practical and ends up being a source of decoherence (the very thing this paper is supposed to avoid) 2) the scheme depends for its gates and construction of a graph state (a kind of resource quantum state) on precise settings of polarized beam splitters, something which will certainly have errors and hence...decoherence again. 3) the author shows how to create a graph state but makes no mention of why this state will persist. Why won't it deco here while waiting for the necessarily adaptive measurement?

http://arxiv.org/abs/1111.1319 for the pre print version of this pay walled publically funded research.

“If a quantum computer existed now, we could solve problems that are exceptionally difficult on current computers, such as cracking codes underlying Internet transactions.”

At least he's being up-front about it, but I'm a little concerned that this consistently seems to be the most prominent projected use of quantum computers. What else are they good for, besides allowing exceptionally wealthy actors to bypass our privacy protections?

They are also good for allowing exceptionally wealthy actors to protect their privacy against other exceptionally wealthy actors who use quantum computers.

Quantum computing is not the same thing as quantum key distribution. QKD is already in use "in the field". QC is still pretty useless and only exists in labs.

There are two flagship applications for quantum computing: (1)Shor's polynomial-time factoring algorithm and (2)Grover's unstructured search algorithm, which has O(√n) running time [classically, if you have an unordered array of n elements fining some element is O(n) so it is a √-speedup].

Both of the above are //theoretical// applications, but the letter agencies seem to be confused about this and think they can actually computers that carry these out. All the quantum researchers working for the US military deserve a Nobel peace prize for all the money they are diverting from actual weapons and shit.

My opinion is that (3)quantum simulations of physical systems will be the most useful part of the quantum computing industry. Also, quantum optics is very cool, and could lead to higher throughput for optical communications.

Other uses:

Quantum simulations. Sampling from solutions to linear equations. Shit we haven't discovered yet (you be shocked at how little research there is in quantum computing looking for new algorithms)

D-Wave machines are good at Travelling Salesman, and it's been suggested to use them for laying out PCBs.

http://www.dwavesys.com/en/dev-tutorial-tsp.html

Reply on news.ycombinator.com