Skip to content

Commit db85850

Browse files
committed
sidebar on post quantum crypto
1 parent cae6c90 commit db85850

File tree

1 file changed

+53
-0
lines changed

1 file changed

+53
-0
lines changed

crypto.rst

Lines changed: 53 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -396,6 +396,59 @@ slower than secret-key ciphers. Consequently, secret-key ciphers are
396396
used for the vast majority of encryption, while public-key ciphers are
397397
reserved for use in authentication and session key establishment.
398398

399+
.. admonition:: Post-Quantum Cryptography
400+
401+
As we have seen, a lot of cryptography depends on the difficulty of
402+
solving certain mathematical problems, such as factoring prime
403+
numbers or computing discrete logarithms. When the efforts of
404+
mathematicians over decades to solve a problem have proven
405+
fruitless, it is tempting to declare these problems sufficiently
406+
hard for our purposes. However, there is a potential weakness
407+
lurking on the horizon, which is that many of these problems are
408+
known to have efficient solutions using quantum computers. Or more
409+
accurately, they could be efficiently solved on quantum computers
410+
that are much larger than any that have been built to date. As
411+
progress is made towards ever larger quantum computers, measured by
412+
the number of quantum bits (qubits), there is a real
413+
risk that many current cryptographic algorithms will at some point
414+
become breakable.
415+
416+
There is plenty of debate about whether quantum computing will ever
417+
progress to the point that the risks to conventional cryptography
418+
materialize. Current quantum computers are much too small and lack
419+
the error-correcting capabilities necessary to solve the
420+
mathematical problems at sufficient scale, and it is not guaranteed
421+
that some version of Moore's law will apply to quantum
422+
computing. Building quantum computers that are large enough (in
423+
number of qubits) and sufficiently fault-tolerant to actually
424+
present a threat to cryptography remains an engineering
425+
challenge. That said, the risk is viewed as being sufficiently
426+
large that steps need to be taken to prepare for the day when
427+
quantum computers *can* break most existing algorithms. It is worth
428+
considering the possibility that some data that is well protected
429+
today could be stored for a decade or two and then decrypted by a
430+
future quantum computer, so even data produced today could be at
431+
risk.
432+
433+
The response to this uncertain threat has been to develop suites of
434+
cryptographic algorithms for which no quantum solution is
435+
known. This is the field of "Post-Quantum Cryptography". Note the
436+
use of the phrase "no solution is known". It is hard to prove that
437+
no algorithm exists—once again we are in the territory of trying to
438+
prove a negative. But NIST is running a process to evaluate and
439+
standardize a set of quantum-resistant algorithms, and there is
440+
plenty of focus on the candidate algorithms to establish their
441+
suitability over the long term.
442+
443+
There is a general, if not universal, sense that at some point
444+
post-quantum cryptographic algorithms will be needed. While the
445+
timeframe is uncertain and the exact algorithms to be used may
446+
change, the requirement for *crypto-agility*—the ability to swap
447+
out one set of algorithms for another—is now well established.
448+
449+
450+
451+
399452
3.4 Message Authentication
400453
---------------------------------
401454

0 commit comments

Comments
 (0)