Re: Wow, care to back that last claim up Simon? :)
Yeah, I'm not thinking offhand of an algorithm in BQP which would be useful here. This isn't my area, but "QC will not help with this problem" is often a safe assumption.
A PCI QC add-on card, even if it provides, say, a hundred reliable effective qubits ("effective" because you will almost certainly need a great many more for error correction), will still have pretty limited applications. Small quantum simulations. Cryptographically-secure random bits (if you can figure out a legitimate real-world application for them1). Breaking small encryption keys and hashes – but 100 qubits is too small for anything in common use that can't already be brute-forced with reasonable resources.
1Arguably there are a number of applications for crypto-secure randomness, such as cryptocurrency, but you have to convince other people that they need it too.