What is the strength of algorithms such as DES (Rijndael) and Blowfish used in password manager Safe password, with the current development (2018) of quantum computing, the former parallel processing, the mandating backdoor in the face of a brute-force attack.

  • 2
    Rijndael is AES not DES. Anyone still using DES or Blowfish is incompetent. – AndrolGenhald Apr 21 '18 at 2:30
  • @AndrolGenhald Blowfish is still useful in certain situations as long as you mind the small block size. – forest Apr 21 '18 at 13:29

DES, especially single DES, cannot be considered secure. A brute force attack in 1998 broke a DES key in 56 hours with a $250,000 specialized machine. More modern estimates are around $10,000 and 24-26 hours to break a DES key, using off-the-shelf tools and without quantum computing. That's not what anyone would consider secure.

Although Blowfish is not as insecure, some attacks can crack messages if few GBs of data have been encrypted using the same key.

"Mandatory backdoors" could break any algorithm.

Although some people suspect that one (or more) such backdoor was inserted into DES, there's no proof. In fact, there is some evidence to the contrary, as some modifications of DES imposed by the NSA actually hardened it against attacks that weren't publicly known.

As for Blowfish, I'm not aware of any claim, let alone evidence, about the existence of a backdoor.

With quantum computing one could brute-force a key with a time complexity that is the square root of the one required by a classical computer. An M-qubit quantum computer, where M is O(log2(2^56)), could break a DES key in 2^28 attempts rather than 2^56.

Currently, AES-256 is considered secure even with quantum computing, while AES-128 is not. DES is not, due to its very limited (56 bit) key size. As for Blowfish, the current default implementation uses 128 bits, which wouldn't be enough to withstand quantum attacks.

  • I think a whole lot more than 56 qubits (I assume you meant that rather than bits) would be required to attack a 56-bit key. – forest Apr 21 '18 at 13:30
  • You are correct, I meant qubits. However, while 56 qubits may not be enough, the order of magnitude wasn't that wrong. Indeed, Grover algorithm (which is at the basis of brute force attacks with quantum computers) requires O(log(N)) qubits, where N is the key space (in our case, 2^56), see these slides by a famous Swiss university. I will edit my answer. – A. Darwin Apr 21 '18 at 15:54

So Blowfish and Des are in current implementation are insecure (see a related article here). Without a doubt this would only get worse as quantum computing emerges. I would suggest going with AES and SHA-512, however even this may not be a challenge for quantum computers due to their insane increase of clock cycles.

  • 1
    Quantum computers do not compute more quickly by having a faster clock rate, but by running algorithms that require prohibitive space-time trade-offs on classical computers. Even still, grover's algorithm only cuts the equivalent bit security of a symmetric cipher in half. – forest Apr 21 '18 at 13:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.