Reply to post: Re: Enquiring minds and all…

Re: Enquiring minds and all…

They *are* digital computers.

It's a hybrid solution, your problem goes to a switch that hands it to a cloud of digital solvers or a DWave box. Whichever is more appropriate.

Quantum computers take hours to setup, cool to superconduct, warm up again. They're finnicky and may need multiple runs. They give answers in hours or days. You would run them only out of curiosity, to see if it can obtain a more optimal solution than digital optimizers (if they actually worked the way Feyman imagined a QM computer would work, they'd always return the most optimal result, sadly none of that shit is, or was ever, real. And they do return sub-optimal results).

Even if you planned on running it on the Quantum box, for marketing purposes, you'd still run the digital solver, get an answer in seconds or minutes just in case the answer is a better solution. The last thing you'd ever want to do, is return an answer to a customer, where the customers own optimizer returns a more optimal answer. If that happened, your game is up, your marketing BS is revealed! Better to also run it on all stock digital solvers as-well to avoid said issue, and return the most optimal of all of them.

I hate this scammy crap, it annoys the hell out of me.

Why not just sell the cloud digital solver? So the customers knows you're delivering the best price, max scalable, best algorithms of the day. And not, say, intentionally running a crap algo like "simulated annealing" just so you can sell them a bit of overpriced hardware that runs simulated annealing faster. A risk with a company that wants to sell something other than the product it's selling.

If you don't understand the main use case for a Quantum Computer its this:

A solver is used to solve problems of the form:

X = Funct(A,B,C,D,E,F,G,H,I,K......)

Find the value of A, B,C,D... for which the function Funct returns the minimum (or maximum) value.

For complex, non-linear systems, often systems derived from real world data, you cannot solve these algebraically. Simply trying brute force limits for each of A, then B, then C, etc. rarely works. Testing all possible values is not possible, there are just too combinations. In addition there may be constraints, e.g. C may only be valid in the range 0 to 1. Worse, when you get to my stuff, you get complex constraints too. e.g. The limits of C depends is some function of E and F.

Solving these is non-trivial, and an entire branch of Math and algorithms.

You could imagine a quantum computer (if it actually worked like Feyman imagined) could solve public key encryption this way.

e.g. X = GrammarCheck(DecryptFunct(EncryptedMessage, A,B,C,D,E,F,G.....))

Find the values of A,B,C,D... that returns the max X, so run a decryptor to find the private keys (A,B,C,D...) that decodes an encrypted message to grammar correct readable text.

You could even test it if you have a test sample and know the result.

e.g. X = CompareForSimilarity(KnownRealDecrypted, DecryptResult(EncryptedMessage, A,B,C,D,E,F,G.....)))

Find the A,B,C,D.... for which X is the minimum, i.e. closest to your known text, then check the A,B,C,D... keys against your known keys (because you're doing a test and did the encryption) and see if it actually found the result.

BUT THIS DOES NOT WORK, it does not work because the core basis for a Quantum Computer is bogus, bunkum, a faulty piece of Science. An approximation model of a system, not the actual system.