Google Achieves Quantum Supremacy. Is Encryption Safe?
tomshardware.com
tomshardware.com
Grover's algorithm only gives a sqrt(n) speedup for breaking AES, so the number of bits is irrelevant; breaking 256-bit AES on a quantum computer would take as much time (or gates, rather) as breaking 128-bit AES on a classical computer.
Scott's Quantum Supremacy FAQ
I wonder if that won't kill some QC investment.
Is it another careful crafted academic lie or is it real this time?
Sure, it's better at some specific tasks. It's not a spreadsheet, let along a general purpose theorem prover, yet.
If we had the quantum equivalent of a 14 button calculator that would be a big deal.
"Quantum supremacy" = "This circuit runs faster than it possibly could have under classical rules, no matter how clever it might be," aka. an asymptotic speedup.
It proves that quantum computation is possible, and the universe isn't just conspiring to pretend quantum mechanics works while actually following something like Copenhagen, but few people were doubting that in the first place.
* quantum supremacy: this is a scientific goal. This can be achieved by building a quantum device that can solve some, possibly useless, computational problem faster than a classical computer. The Google team claims to have done so. For scientists, wondering whether the extended Church-Turing thesis is true or not, the Google result is a significant result. It provides some evidence that the thesis is not true.
* quantum advantage: this is a technological goal (even though it also has scientific implications). This is about building a quantum device that actually solves useful computational problems (like breaking encryption) faster than a classical computer. We are not there yet.
Its like e.g. sending a sounding rocket into space and back down. It doesn't reach orbit or carry a useful payload, but it shows that a rocket could probably reach orbit. Not a perfect analogy though since in the rocket case it's really a matter of just scaling the rocket while for QC there remain large known unknowns and probably many unknown unknowns.
---
1) deltaV - the change in velocity you need to reach your destination. this is completely fixed. for instance if you want to get into earth's orbit you need to be able to change your velocity by 8km/s regardless of any other variable.
2) specific impulse of your rocket - again this is mostly fixed. it can be improved by technological breakthroughs, but not by things like increasing scale.
3) fuel to mass ratio - what percent of your rocket is made up of fuel
---
Any two define the third. deltaV constrains absolutely everything we can do, so we can set it as the goal. Specific impulse is not something that can be readily improved for the most part, so all you're left with is a fuel to mass ratio. And suffice to say making things bigger while making them hold an even larger ratio of fuel is not how things work. Also important is that based on our requirements for the first two variables, our effective fuel to mass ratio is already approaching very close to 100%.
We of course take it all for granted now and SpaceX going up and softly autonomously landing back down, sometimes in the middle of the ocean on an autonomous barge, is increasingly starting to look almost mundane. But it's borderline miraculous we can achieve this, and indeed if Earth was just slightly larger (about a 50% greater diameter - just some 4,000 miles) it wouldn't even be possible for us to leave this planet with anything like current technology, regardless of how epic a scale we built upon.
[1] - https://www.nasa.gov/mission_pages/station/expeditions/exped...
I'd take out that parenthesis, it's misleading. The Church-Turing thesis is about computability, whereas the "extended" one is about complexity. They're quite different beasts despite the similar name, and QS would really only be evidence against the extended version.