Super cool.
If current quantum computers were scaled up to more qubits, could they break modern crypto? Or would we need both more qubits and a new quantum computer architecture?
If current quantum computers were scaled up to more qubits, could they break modern crypto? Or would we need both more qubits and a new quantum computer architecture?
I wonder how possible it is that IBM could have already gone further and are already cracking modern crypto in secret, e.g. funded by the NSA. Is that a crazy conspiracy idea, or actually a possibility?
I am investing in IBM under the assumption that this is an actual possibility. Their public QC roadmap actually looks like a realistic journey now.
I strongly believe that the NSA, et. al. currently have access to a very powerful quantum computer - likely constructed by IBM under contract.
The game theory around this is such that it is impossible for me to accept that there are zero secret quantum computers in existence by now. There is too much to lose by not playing that game as hard as you can.
(I deleted my previous edit because I had misread part of what you wrote.)
The challenge of building QCs - as evidenced by billions of dollars worth of research in them - is many orders of magnitude more difficult than say the Manhattan project. The latter put together the best of the best on the project. You are suggesting a scenario where a tiny fraction of the best of the best are secreted away, with many of their past collaborators unaware of their doings, and have successfully built a QC.
While the many brilliant best of the best who are working publicly, with many billions of dollars of research funding are currently only making very slow progress. It simply does not square.
For example their Accumulo OSS suggests they're capturing and storing a lot of data to analyze later. The Ghidra OSS being a best in class reverse engineering tool also suggests that alot of their work revolves around finding zero day vulnerabilities.
I would guess they probably don't have any actually useful and in production right now, but they probably have a few secreted away in development, so they will be ready to put them to use if/when they do become useful.
That depends on what you mean by "scaled up". There is a concept of "Quantum Volume" that exists, which basically means the depth of the longest qubit circuit you can pull off.
https://en.wikipedia.org/wiki/Quantum_volume
'Simply' (it's never simple ;) ) adding qubits to a machine does not necessarily increase its Quantum Volume. Decreasing the noise typically will.
However, there is a threshold at which point you can scale up mostly indefinitely. This is what the whole Quantum Error Correction is all about.
https://en.wikipedia.org/wiki/Quantum_error_correction
There is a paper
https://arxiv.org/abs/1905.09749
That goes into a clear discussion of how to build a quantum computer and the associated thresholds that would allow you to do so. There is a minimum number of qubits needed (that work perfectly), but the paper analyzes how many qubits you'd need under realistic assumptions about how many noisy qubits you'd need to get error correcting qubits at the needed reliability.