Quantum Computers Are the Future Nukes of the IT World

https://www.nextplatform.com/2018/09/20/quantum-computers-are-the-future-nukes-of-the-it-world/

by rbanffy • 8 years ago

20 33 8 years ago

ohiovr7• 8 years ago

Are ordinary ssh keys vulnerable?

Can you elaborate more? In what sense? They are certainly vulnerable if you compromise the system which has them saved. Are they vulnerable to a brute force breach? Likely not, brute force is always a slow process, no matter how you speed it up. Is X vulnerable should always be phrased as Is X vulnerable to Y.

ohiovr7• 8 years ago

right, brute force your way in without a key

archgoon3• 8 years ago

If your ssh key relies on RSA, then yes, you're relying on the difficulty of factoring numbers; which Shor's algorithm can perform efficiently. But only if the quantum computer has enough qubits. The needed number of qubits needed to factor is a multiple of the rsa key length. Unlike a regular computer, where you can finagle two computers with 2 gigs of memory to work like a single machine wiht 4 gigabytes of memory, so you can't just merge two 1024 qubit computers to get a 2048 qubit computer. And we're a long way off to having anything close to that.

The simplest solution is simply to increase the size of the key beyond that of the most advanced quantum computer (which is essentially today's approach ;) ). Also, you just need to exchange a symmetric key; Shor's algorithm is a special case, you don't get the exponential -> polynomial speedup[1] to encryption algorithms in general. So once you deliver the symmetric key in a secure way, using a quantum resistant algorithm, you can resume business as usual.

[1] Though in brute force searches, you can reduce the effective key length by 2. https://en.wikipedia.org/wiki/Grover%27s_algorithm

ohiovr7• 8 years ago

You can't have an an infinitely long key. Eventually the decripting process slows everything down to nothing.

hannob3• 8 years ago

Short and simple: yes.

Though the question is how relevant this is. Likely SSH will have moved to something more secure years before any quantum computer with sufficient computing power will become available.

Because that's the thing these scare-stories don't tell you: While the progress in quantum computing is moving very slowly, in recent years the progress in post quantum cryptography has been very fast.

ohiovr7• 8 years ago

A quantum computer with threatinging amounts of qubits could already exist as a secret military project today.

domparise2• 8 years ago

Quantum fintech is going to make blockchain fintech look like child’s play.

madeuptempacct• 8 years ago

You can't make statements like that without elaborating on what quantum computing can possibly provide.

domparise2• 8 years ago

Two examples off the top of my head: 1. Instantaneous stock transactions from anywhere in the world. (quantum communication/entanglement) Offers a substantial RTT improvement on current trading systems.

2. High frequency trading based on quantum prediction algorithms. (Quantum information theory) May provide for HFT algorithms capable of producing predictions based on an exponential number of information streams. E.g. imagine making future investment decisions based on _all_ trades that happen in a span of time.

@ me if you decide make a company to do either of these things.

gl00pp• 8 years ago

I'd have to say that the blockchain would just go quantum too. NBD

expertentipp• 8 years ago

We can expect a flood of quantum fintech machine learning PhDs in couple of years.

lostmsu• 8 years ago

I think you accidentally missed the word "blockchain" somewhere.

GenericsMotors• 8 years ago

Don't need quantum fintech for that, "blockchain fintech" is already rubbish.

amelius2• 8 years ago

Question: I get that QC can break some cryptosystems, but can it break all known cryptosystems?

detaro• 8 years ago

Symmetric encryption generally is ok/only weakened, and there's lots of research into solutions: https://en.wikipedia.org/wiki/Post-quantum_cryptography

archgoon3• 8 years ago

To our knowledge, it cannot. It would be extremely surprising if it could.

However, it can reduce the difficulty of a generic brute force search

https://en.wikipedia.org/wiki/Grover%27s_algorithm

But it only reduces the effective key length by 2. Which, although amazing, has a fairly straightforward response. Double your key length beyond what was previously viewed as 'intractable'.

yters2• 8 years ago

Nothing can break a one time pad, used for the super secure stuff.

amelius2• 8 years ago

Yes, good point.

hannob3• 8 years ago

They break all widely used public key encryption systems (RSA, DH, ECC-based). There are post-quantum systems, but none of them are in widespread use yet.

Symmetric encryption systems are weakened, but can easily be made quantum safe by increasing the key size.

eximius• 8 years ago

No, they aren't.

You can't remotely disarm nuclear weapons or protect an entire industry by switching algorithms.

They are a problem, and we should prepare for them, but they aren't apocalyptic.

archgoon3• 8 years ago

Unlike nuclear weapons, a quantum computer is only dangerous if people don't know you have one.

There are a number of ways of dealing with the security issues posed by quantum computers; most of them involve simply increasing the key length of your encryption keys.

However, if people don't think you have a quantum computer, or you're able to make it better faster than expected, they won't change their behavior and so you have a fixed target.

yters2• 8 years ago

It's not clear that quantum computers are even all that magical. At least we know they are no more powerful than a Turing machine.

hannob3• 8 years ago

Second sentence: "Now that the technology is moving from research to practice"

No such thing is happening. We're still in the state of "we're waiting for someone to actually show that quantum computers can do anything at all faster than a classical computer". Yeah, it's probably gonna happen in the not so far future, yeah, there's progress. But saying anything is moving "to practice" is absurd at this point.

08-15• 8 years ago

This is not news, it's an advertisement.

It's all fluff, vague nonsense about what quantum computers might supposedly do, and mostly wrong. The implied assumption is that a quantum computer can solve any NP problem in polynomial time, and that doesn't seem likely. And then, all of a sudden, the purpose of this nonsense becomes clear:

For its part, ISARA has developed a new method of creating a compound or hybrid digital certificate

This is followed by lots of marketing buzzwords. Disgusting, really.

Quantum Computers Are the Future Nukes of the IT World

Related Stories